Journal articles on the topic 'Novel task transfer'
To see the other types of publications on this topic, follow the link: Novel task transfer.
Create a spot-on reference in APA, MLA, Chicago, Harvard, and other styles
Consult the top 50 journal articles for your research on the topic 'Novel task transfer.'
Next to every source in the list of references, there is an 'Add to bibliography' button. Press on it, and we will generate automatically the bibliographic reference to the chosen work in the citation style you need: APA, MLA, Harvard, Chicago, Vancouver, etc.
You can also download the full text of the academic publication as pdf and read online its abstract whenever available in the metadata.
Browse journal articles on a wide variety of disciplines and organise your bibliography correctly.
1
Vincent, Vercruyssen, Meert Wannes, and Davis Jesse. "Transfer Learning for Anomaly Detection through Localized and Unsupervised Instance Selection." Proceedings of the AAAI Conference on Artificial Intelligence 34, no. 04 (April 3, 2020): 6054–61. http://dx.doi.org/10.1609/aaai.v34i04.6068.
Full textAbstract:
Anomaly detection attempts to identify instances that deviate from expected behavior. Constructing performant anomaly detectors on real-world problems often requires some labeled data, which can be difficult and costly to obtain. However, often one considers multiple, related anomaly detection tasks. Therefore, it may be possible to transfer labeled instances from a related anomaly detection task to the problem at hand. This paper proposes a novel transfer learning algorithm for anomaly detection that selects and transfers relevant labeled instances from a source anomaly detection task to a target one. Then, it classifies target instances using a novel semi-supervised nearest-neighbors technique that considers both unlabeled target and transferred, labeled source instances. The algorithm outperforms a multitude of state-of-the-art transfer learning methods and unsupervised anomaly detection methods on a large benchmark. Furthermore, it outperforms its rivals on a real-world task of detecting anomalous water usage in retail stores.
2
Kearney, Philip E., and Phil Judge. "Successful Transfer of a Motor Learning Strategy to a Novel Sport." Perceptual and Motor Skills 124, no. 5 (July 7, 2017): 1009–21. http://dx.doi.org/10.1177/0031512517719189.
Full textAbstract:
This study investigated whether secondary school students who were taught a motor learning strategy could transfer their knowledge of the strategy to learning a novel task. Twenty adolescents were randomly allocated to a strategy or control group. The strategy group was taught Singer’s five-step learning strategy, while the control group received information on the evolution and biomechanics of the basketball free throw. Both groups received three 1-hour practice sessions on a modified basketball shooting task. After one month, participants were introduced to the transfer task, golf putting. Performance accuracy was recorded for all tasks, and participants completed questionnaires regarding strategy use during practice. Participants taught the five-step learning strategy successfully recalled and applied it after a 1-month interval, and they demonstrated superior performance on both acquisition and transfer tasks, relative to the control group. Physical education teachers and coaches should consider using this learning strategy to enhance the learning of closed motor skills.
3
Catrambone, Richard. "Specific versus General Instructions: Initial Performance and Later Transfer." Proceedings of the Human Factors Society Annual Meeting 33, no. 19 (October 1989): 1320–23. http://dx.doi.org/10.1177/154193128903301918.
Full textAbstract:
Two experiments demonstrated that people who receive specific instructions (SI subjects) for using a word processor are able to accomplish initial tasks more quickly than people who receive more general instructions (GI subjects). Experiment 1 found, however, that SI subjects were unable to do a novel transfer task unless they received hints while GI subjects had no trouble with the transfer task. A production rule analysis was used to guide a revision of the specific instructions so that those instructions promoted generalization. Experiment 2 used these revised specific instructions and found that SI subjects were now able to do a novel transfer task about as well as GI subjects. These results suggest that a production system is a useful tool for analyzing instructions and predicting user performance and that specific instructions designed to promote generalization may be the most effective type of instructions.
4
GARCÍA, ESTEBAN O., ENRIQUE MUNOZ DE COTE, and EDUARDO F. MORALES. "TRANSFER LEARNING FOR CONTINUOUS STATE AND ACTION SPACES." International Journal of Pattern Recognition and Artificial Intelligence 28, no. 07 (October 14, 2014): 1460007. http://dx.doi.org/10.1142/s0218001414600076.
Full textAbstract:
Transfer learning focuses on developing methods to reuse information gathered from a source task in order to improve the learning performance in a related task. In this work, we present a novel approach to transfer knowledge between tasks in a reinforcement learning (RL) framework with continuous states and actions, where the transition and policy functions are approximated by Gaussian processes. The novelty in the proposed approach lies in the idea of transferring information about the hyper-parameters of the state transition function from the source task, which represents qualitative knowledge about the type of transition function that the target task might have, constraining the search space and accelerating the learning process. We performed experiments on relevant tasks for RL, which show a clear improvement in the overall performance when compared to state-of-the-art reinforcement learning and transfer learning algorithms for continuous state and action spaces.
5
Kono, Hitoshi, Yuto Sakamoto, Yonghoon Ji, and Hiromitsu Fujii. "Automatic Transfer Rate Adjustment for Transfer Reinforcement Learning." International Journal of Artificial Intelligence & Applications 11, no. 6 (November 30, 2020): 47–54. http://dx.doi.org/10.5121/ijaia.2020.11605.
Full textAbstract:
This paper proposes a novel parameter for transfer reinforcement learning to avoid over-fitting when an agent uses a transferred policy from a source task. Learning robot systems have recently been studied for many applications, such as home robots, communication robots, and warehouse robots. However, if the agent reuses the knowledge that has been sufficiently learned in the source task, deadlock may occur and appropriate transfer learning may not be realized. In the previous work, a parameter called transfer rate was proposed to adjust the ratio of transfer, and its contribution include avoiding dead lock in the target task. However, adjusting the parameter depends on human intuition and experiences. Furthermore, the method for deciding transfer rate has not discussed. Therefore, an automatic method for adjusting the transfer rate is proposed in this paper using a sigmoid function. Further, computer simulations are used to evaluate the effectiveness of the proposed method to improve the environmental adaptation performance in a target task, which refers to the situation of reusing knowledge.
6
Abdelrahman, Amro M., Denny Yu, Bethany R. Lowndes, EeeLN H. Buckarma, Becca L. Gas, David R. Farley, Juliane Bingener, and M. Susan Hallbeck. "Validation of a Novel Inverted Peg Transfer Task: Advancing Beyond the Regular Peg Transfer Task for Surgical Simulation-Based Assessment." Journal of Surgical Education 75, no. 3 (May 2018): 836–43. http://dx.doi.org/10.1016/j.jsurg.2017.09.028.
Full text
7
Lei, Feifei, Jieren Cheng, Yue Yang, Xiangyan Tang, Victor S. Sheng, and Chunzao Huang. "Improving Heterogeneous Network Knowledge Transfer Based on the Principle of Generative Adversarial." Electronics 10, no. 13 (June 24, 2021): 1525. http://dx.doi.org/10.3390/electronics10131525.
Full textAbstract:
Deep learning requires a large amount of datasets to train deep neural network models for specific tasks, and thus training of a new model is a very costly task. Research on transfer networks used to reduce training costs will be the next turning point in deep learning research. The use of source task models to help reduce the training costs of the target task models, especially heterogeneous systems, is a problem we are studying. In order to quickly obtain an excellent target task model driven by the source task model, we propose a novel transfer learning approach. The model linearly transforms the feature mapping of the target domain and increases the weight value for feature matching to realize the knowledge transfer between heterogeneous networks and add a domain discriminator based on the principle of generative adversarial to speed up feature mapping and learning. Most importantly, this paper proposes a new objective function optimization scheme to complete the model training. It successfully combines the generative adversarial network with the weight feature matching method to ensure that the target model learns the most beneficial features from the source domain for its task. Compared with the previous transfer algorithm, our training results are excellent under the same benchmark for image recognition tasks.
8
Jing, Mingxuan, Xiaojian Ma, Wenbing Huang, Fuchun Sun, and Huaping Liu. "Task Transfer by Preference-Based Cost Learning." Proceedings of the AAAI Conference on Artificial Intelligence 33 (July 17, 2019): 2471–78. http://dx.doi.org/10.1609/aaai.v33i01.33012471.
Full textAbstract:
The goal of task transfer in reinforcement learning is migrating the action policy of an agent to the target task from the source task. Given their successes on robotic action planning, current methods mostly rely on two requirements: exactlyrelevant expert demonstrations or the explicitly-coded cost function on target task, both of which, however, are inconvenient to obtain in practice. In this paper, we relax these two strong conditions by developing a novel task transfer framework where the expert preference is applied as a guidance. In particular, we alternate the following two steps: Firstly, letting experts apply pre-defined preference rules to select related expert demonstrates for the target task. Secondly, based on the selection result, we learn the target cost function and trajectory distribution simultaneously via enhanced Adversarial MaxEnt IRL and generate more trajectories by the learned target distribution for the next preference selection. The theoretical analysis on the distribution learning and convergence of the proposed algorithm are provided. Extensive simulations on several benchmarks have been conducted for further verifying the effectiveness of the proposed method.
9
Toader, Andra F., and Thomas Kessler. "Task Variation and Mental Models Divergence Influencing the Transfer of Team Learning." Small Group Research 49, no. 5 (July 26, 2018): 545–75. http://dx.doi.org/10.1177/1046496418786429.
Full textAbstract:
We investigate how teams develop and transfer general problem-solving skills across two ill-structured problems. We draw on cognitive flexibility theory in the instructional literature and propose that teams will achieve a higher performance on a novel task or transfer when they receive an external task intervention (i.e., task variation) and when the internal mechanisms (i.e., divergent mental models) are developed to make sense of the external intervention. To test these predictions, we designed a longitudinal experiment with 17 student teams that encountered task variation during their work on an initial task. Consistent with our predictions, we found that teams that experienced variations and whose mental models diverged during their work on an initial task achieved higher performance on a novel task than teams that experienced variation and whose mental models converged. Implications for the transfer of learning in teams on ill-structured problems are discussed.
10
Vandenbroucke, B., and P. Camps. "CMACIONIZE 2.0: a novel task-based approach to Monte Carlo radiation transfer." Astronomy & Astrophysics 641 (September 2020): A66. http://dx.doi.org/10.1051/0004-6361/202038364.
Full textAbstract:
Context. Monte Carlo radiative transfer (MCRT) is a widely used technique to model the interaction between radiation and a medium. It plays an important role in astrophysical modelling and when these models are compared with observations. Aims. We present a novel approach to MCRT that addresses the challenging memory-access patterns of traditional MCRT algorithms, which prevent an optimal performance of MCRT simulations on modern hardware with a complex memory architecture. Methods. We reformulated the MCRT photon-packet life cycle as a task-based algorithm, whereby the computation is broken down into small tasks that are executed concurrently. Photon packets are stored in intermediate buffers, and tasks propagate photon packets through small parts of the computational domain, moving them from one buffer to another in the process. Results. Using the implementation of the new algorithm in the photoionization MCRT code CMACIONIZE 2.0, we show that the decomposition of the MCRT grid into small parts leads to a significant performance gain during the photon-packet propagation phase, which constitutes the bulk of an MCRT algorithm because memory caches are used more efficiently. Our new algorithm is faster by a factor 2 to 4 than an equivalent traditional algorithm and shows good strong scaling up to 30 threads. We briefly discuss adjustments to our new algorithm and extensions to other astrophysical MCRT applications. Conclusions. We show that optimising the memory access patterns of a memory-bound algorithm such as MCRT can yield significant performance gains.
11
Madhavan, Poornima, Cleotilde Gonzalez, and Frank C. Lacson. "Differential Base Rate Training Influences Detection of Novel Targets in a Complex Visual Inspection Task." Proceedings of the Human Factors and Ergonomics Society Annual Meeting 51, no. 4 (October 2007): 392–96. http://dx.doi.org/10.1177/154193120705100451.
Full textAbstract:
We studied the effects that multiple levels of signal probability (known as base rate) have on the transfer of learning in an airline luggage screening task. Participants (n = 33) were presented with three base rates during the acquisition (training) phase: 100%, 50%, or 20%; at transfer, all participants detected novel targets at a base rate of 20%. Performance was measured by hit rates, false alarm rates, sensitivities, and detection times. Results revealed that participants receiving higher base rates during training obtained higher hit rates at transfer compared to participants encountering lower base rates. However, increasing the training base rate also increased the incidence of false alarms, leading to a low overall level of sensitivity during transfer. Relatively higher base rates had mixed effects on response times. These results have implications for improving training modules for individuals in complex visual inspection tasks.
12
Zhou, Yi, and Fenglei Yang. "Latent Structure Matching for Knowledge Transfer in Reinforcement Learning." Future Internet 12, no. 2 (February 13, 2020): 36. http://dx.doi.org/10.3390/fi12020036.
Full textAbstract:
Reinforcement learning algorithms usually require a large number of empirical samples and give rise to a slow convergence in practical applications. One solution is to introduce transfer learning: Knowledge from well-learned source tasks can be reused to reduce sample request and accelerate the learning of target tasks. However, if an unmatched source task is selected, it will slow down or even disrupt the learning procedure. Therefore, it is very important for knowledge transfer to select appropriate source tasks that have a high degree of matching with target tasks. In this paper, a novel task matching algorithm is proposed to derive the latent structures of value functions of tasks, and align the structures for similarity estimation. Through the latent structure matching, the highly-matched source tasks are selected effectively, from which knowledge is then transferred to give action advice, and improve exploration strategies of the target tasks. Experiments are conducted on the simulated navigation environment and the mountain car environment. The results illustrate the significant performance gain of the improved exploration strategy, compared with traditional ϵ -greedy exploration strategy. A theoretical proof is also given to verify the improvement of the exploration strategy based on latent structure matching.
13
Cao, Yu, and Hua Xu. "SATNet: Symmetric Adversarial Transfer Network Based on Two-Level Alignment Strategy towards Cross-Domain Sentiment Classification (Student Abstract)." Proceedings of the AAAI Conference on Artificial Intelligence 34, no. 10 (April 3, 2020): 13763–64. http://dx.doi.org/10.1609/aaai.v34i10.7153.
Full textAbstract:
In recent years, domain adaptation tasks have attracted much attention, especially, the task of cross-domain sentiment classification (CDSC). In this paper, we propose a novel domain adaptation method called Symmetric Adversarial Transfer Network (SATNet). Experiments on the Amazon reviews dataset demonstrate the effectiveness of SATNet.
14
Al-Smadi, Mohammed, Mahmoud Hammad, Qanita Bani Baker, Saja Khaled Tawalbeh, and Sa’ad A. Al-Zboon. "Transfer deep learning approach for detecting coronavirus disease in X-ray images." International Journal of Electrical and Computer Engineering (IJECE) 11, no. 6 (December 1, 2021): 4999. http://dx.doi.org/10.11591/ijece.v11i6.pp4999-5008.
Full textAbstract:
<p><span lang="EN-US">Currently, the whole world is fighting a very dangerous and infectious disease caused by the novel coronavirus, called COVID-19. The COVID-19 is rapidly spreading around the world due to its high infection rate. Therefore, early discovery of COVID-19 is crucial to better treat the infected person as well as to slow down the spread of this virus. However, the current solution for detecting COVID-19 cases including the PCR test, CT images, epidemiologically history, and clinical symptoms suffer from high false positive. To overcome this problem, we have developed a novel transfer deep learning approach for detecting COVID-19 based on x-ray images. Our approach helps medical staff in determining if a patient is normal, has COVID-19, or other pneumonia. Our approach relies on pre-trained models including Inception-V3, Xception, and MobileNet to perform two tasks: i) binary classification to determine if a person infected with COVID-19 or not and ii) a multi-task classification problem to distinguish normal, COVID-19, and pneumonia cases. Our experimental results on a large dataset show that the F1-score is 100% in the first task and 97.66 in the second task.</span></p>
15
Ahmed, Alaa A., and Daniel M. Wolpert. "Transfer of Dynamic Learning Across Postures." Journal of Neurophysiology 102, no. 5 (November 2009): 2816–24. http://dx.doi.org/10.1152/jn.00532.2009.
Full textAbstract:
When learning a difficult motor task, we often decompose the task so that the control of individual body segments is practiced in isolation. But on re-composition, the combined movements can result in novel and possibly complex internal forces between the body segments that were not experienced (or did not need to be compensated for) during isolated practice. Here we investigate whether dynamics learned in isolation by one part of the body can be used by other parts of the body to immediately predict and compensate for novel forces between body segments. Subjects reached to targets while holding the handle of a robotic, force-generating manipulandum. One group of subjects was initially exposed to the novel robot dynamics while seated and was then tested in a standing position. A second group was tested in the reverse order: standing then sitting. Both groups adapted their arm dynamics to the novel environment, and this movement learning transferred between seated and standing postures and vice versa. Both groups also generated anticipatory postural adjustments when standing and exposed to the force field for several trials. In the group that had learned the dynamics while seated, the appropriate postural adjustments were observed on the very first reach on standing. These results suggest that the CNS can immediately anticipate the effect of learned movement dynamics on a novel whole-body posture. The results support the existence of separate mappings for posture and movement, which encode similar dynamics but can be adapted independently.
16
Stanford, Sophia, Andrew Milne, and Jennifer MacRitchie. "The Effect of Isomorphic Pitch Layouts on the Transfer of Musical Learning †." Applied Sciences 8, no. 12 (December 6, 2018): 2514. http://dx.doi.org/10.3390/app8122514.
Full textAbstract:
The physical arrangement of pitches in most traditional musical instruments—including the piano and guitar—is non-isomorphic, which means that a given spatial relationship between two keys, buttons, or fretted strings can produce differing musical pitch intervals. Recently, a number of new musical interfaces have been developed with isomorphic pitch layouts where these relationships are consistent. Since the nineteenth century, it has been widely considered that isomorphic pitch layouts facilitate the learnability and playability of instruments, particularly when a piece is transposed into a different key; however, prior to this paper, this has not been experimentally tested. To address this, we investigated four different pitch layouts to examine whether isomorphism facilitates retention and transfer of musical learning within and across keys. Both non-musicians and musicians were tested on two training tasks: two immediate retention tasks and a transfer task. Each participant played every task on two distinct layouts—one being an isomorphic layout (Wicki or Bosanquet), the other being a minimally adjusted non-isomorphic version. For musicians, isomorphism was found to facilitate transfer of learning to a novel task; for non-musicians, the results were mixed. This study provides insight into features that are important to music instrument design.
17
Lan, Wu, and Han Xiaolei Han. "Novel Steganalysis Method for Unknown Embedding Rates using Transfer and Multi-Task Learning." International Journal of Performability Engineering 15, no. 12 (2019): 3139. http://dx.doi.org/10.23940/ijpe.19.12.p5.31393150.
Full text
18
Ding, Guanwen, Yubin Liu, Xizhe Zang, Xuehe Zhang, Gangfeng Liu, and Jie Zhao. "A Task-Learning Strategy for Robotic Assembly Tasks from Human Demonstrations." Sensors 20, no. 19 (September 25, 2020): 5505. http://dx.doi.org/10.3390/s20195505.
Full textAbstract:
In manufacturing, traditional task pre-programming methods limit the efficiency of human–robot skill transfer. This paper proposes a novel task-learning strategy, enabling robots to learn skills from human demonstrations flexibly and generalize skills under new task situations. Specifically, we establish a markerless vision capture system to acquire continuous human hand movements and develop a threshold-based heuristic segmentation algorithm to segment the complete movements into different movement primitives (MPs) which encode human hand movements with task-oriented models. For movement primitive learning, we adopt a Gaussian mixture model and Gaussian mixture regression (GMM-GMR) to extract the optimal trajectory encapsulating sufficient human features and utilize dynamical movement primitives (DMPs) to learn for trajectory generalization. In addition, we propose an improved visuo-spatial skill learning (VSL) algorithm to learn goal configurations concerning spatial relationships between task-relevant objects. Only one multioperation demonstration is required for learning, and robots can generalize goal configurations under new task situations following the task execution order from demonstration. A series of peg-in-hole experiments demonstrate that the proposed task-learning strategy can obtain exact pick-and-place points and generate smooth human-like trajectories, verifying the effectiveness of the proposed strategy.
19
Alzubaidi, Laith, Muthana Al-Amidie, Ahmed Al-Asadi, Amjad J. Humaidi, Omran Al-Shamma, Mohammed A. Fadhel, Jinglan Zhang, J. Santamaría, and Ye Duan. "Novel Transfer Learning Approach for Medical Imaging with Limited Labeled Data." Cancers 13, no. 7 (March 30, 2021): 1590. http://dx.doi.org/10.3390/cancers13071590.
Full textAbstract:
Deep learning requires a large amount of data to perform well. However, the field of medical image analysis suffers from a lack of sufficient data for training deep learning models. Moreover, medical images require manual labeling, usually provided by human annotators coming from various backgrounds. More importantly, the annotation process is time-consuming, expensive, and prone to errors. Transfer learning was introduced to reduce the need for the annotation process by transferring the deep learning models with knowledge from a previous task and then by fine-tuning them on a relatively small dataset of the current task. Most of the methods of medical image classification employ transfer learning from pretrained models, e.g., ImageNet, which has been proven to be ineffective. This is due to the mismatch in learned features between the natural image, e.g., ImageNet, and medical images. Additionally, it results in the utilization of deeply elaborated models. In this paper, we propose a novel transfer learning approach to overcome the previous drawbacks by means of training the deep learning model on large unlabeled medical image datasets and by next transferring the knowledge to train the deep learning model on the small amount of labeled medical images. Additionally, we propose a new deep convolutional neural network (DCNN) model that combines recent advancements in the field. We conducted several experiments on two challenging medical imaging scenarios dealing with skin and breast cancer classification tasks. According to the reported results, it has been empirically proven that the proposed approach can significantly improve the performance of both classification scenarios. In terms of skin cancer, the proposed model achieved an F1-score value of 89.09% when trained from scratch and 98.53% with the proposed approach. Secondly, it achieved an accuracy value of 85.29% and 97.51%, respectively, when trained from scratch and using the proposed approach in the case of the breast cancer scenario. Finally, we concluded that our method can possibly be applied to many medical imaging problems in which a substantial amount of unlabeled image data is available and the labeled image data is limited. Moreover, it can be utilized to improve the performance of medical imaging tasks in the same domain. To do so, we used the pretrained skin cancer model to train on feet skin to classify them into two classes—either normal or abnormal (diabetic foot ulcer (DFU)). It achieved an F1-score value of 86.0% when trained from scratch, 96.25% using transfer learning, and 99.25% using double-transfer learning.
20
Resing, Wilma C. M., Erika Tunteler, and Julian G. Elliott. "The Effect of Dynamic Testing With Electronic Prompts and Scaffolds on Children’s Inductive Reasoning: A Microgenetic Study." Journal of Cognitive Education and Psychology 14, no. 2 (2015): 231–51. http://dx.doi.org/10.1891/1945-8959.14.2.231.
Full textAbstract:
The study aimed to gain insight into children’s variability in inductive reasoning problem-solving strategies. Visual–spatial series completion tasks were employed that drew on the novel use of electronic tangibles. Two approaches were contrasted: repeated practice alone and repeated practice with the addition of dynamic training. Participants were 38 children aged 6–9 years, who were allocated to 1 of 2 treatment conditions. In condition one, children had opportunities for repeated practice on 4 sessions. The children in the second condition received the same repeated practice sessions but, in addition, were also provided with training. Transfer of learning was measured before and after the practice and training sessions. During the sessions, the children were presented with series completion tasks using tangible objects, each time with a 1–week interval. In comparison with the repeated practice alone condition, the children with additional training showed significantly greater gains in performance (accuracy and efficiency). Findings clearly showed inter- and intravariability in children’s use of problem-solving strategies, which decreased after training. There was evidence of transfer of inductive reasoning from the original series completion task (using concrete, discrete elements) to a series completion task with numbers and geometric forms. In summary, this study revealed individual differences and variability in the sorts of help required, (more) stable progression of these results, and the child’s ability to transfer learning to novel tasks and situations.
21
Cao, Huazhen, Tao Yu, Xiaoshun Zhang, Bo Yang, and Yaxiong Wu. "Reactive Power Optimization of Large-Scale Power Systems: A Transfer Bees Optimizer Application." Processes 7, no. 6 (May 31, 2019): 321. http://dx.doi.org/10.3390/pr7060321.
Full textAbstract:
A novel transfer bees optimizer for reactive power optimization in a high-power system was developed in this paper. Q-learning was adopted to construct the learning mode of bees, improving the intelligence of bees through task division and cooperation. Behavior transfer was introduced, and prior knowledge of the source task was used to process the new task according to its similarity to the source task, so as to accelerate the convergence of the transfer bees optimizer. Moreover, the solution space was decomposed into multiple low-dimensional solution spaces via associated state-action chains. The transfer bees optimizer performance of reactive power optimization was assessed, while simulation results showed that the convergence of the proposed algorithm was more stable and faster, and the algorithm was about 4 to 68 times faster than the traditional artificial intelligence algorithms.
22
Wu, Fang Jun. "Empirical Comparative Study of Boosting and its Relatives." Applied Mechanics and Materials 312 (February 2013): 667–72. http://dx.doi.org/10.4028/www.scientific.net/amm.312.667.
Full textAbstract:
Transfer learning is an important research topic in machine learning and data mining that focuses on utilizing knowledge and skills learned in previous tasks to a novel but related task. This paper contributes to comparison between boosting for transfer learning and boosting. The results, in terms of the accuracy, weighted F-Measure, G-Mean, weighted GMPR, weighted precision and weighted AUC, are rigorously tested using the statistical framework proposed by Janez Demsar. Results show that the performance difference between TrAdaBoost and AdaBoost is less significant.
23
Tong, Christine, and J. Randall Flanagan. "Task-Specific Internal Models for Kinematic Transformations." Journal of Neurophysiology 90, no. 2 (August 2003): 578–85. http://dx.doi.org/10.1152/jn.01087.2002.
Full textAbstract:
Numerous studies of motor learning have focused on how people adapt their reaching movements to novel dynamic and visuomotor perturbations that alter the actual or visually perceived motion of the hand. An important finding from this work is that learning of novel dynamics generalizes across different movement tasks. Thus adaptation to an unusual force field generalizes from center-out reaching movements to circular movements ( Conditt et al. 1997 ). This suggests that subjects acquired an internal model of the dynamic environment that could be used to determine the motor commands needed for untrained movements. Using a task interference paradigm, we investigated whether transfer across tasks is also observed when learning visuomotor transformations. On day 1, all subjects adapted to a +30° rotation while making center-out-and-back reaching movements. After a delay of 5 min, different groups of subjects then adapted to a –30° rotation while performing either a continuous tracking task, a figure-eight drawing task, or the center-out-and-back reaching task. All subjects were then retested the next day on the +30° rotation in the reaching task. As expected, subjects who experienced the opposing rotations while performing the same reaching tasks showed no retention of learning for the first rotation when tested on day 2 ( Krakauer et al. 1999 ). In contrast, such retrograde interference was not observed in the two groups of subjects who experienced the opposing rotations while performing different tasks. In fact, their performance on day 2 was similar to that of control subjects who never experienced the opposite rotation. This lack of interference suggests that memory resources for visuomotor rotations are task specific.
24
Byrd, Ronald, and Melvin Gibson. "Bilateral Transfer in Mentally Retarded Children of Ages 7 to 17 Years." Perceptual and Motor Skills 66, no. 1 (February 1988): 115–19. http://dx.doi.org/10.2466/pms.1988.66.1.115.
Full textAbstract:
The purpose of this study was to describe the effect of age on bilateral transfer of mildly mentally retarded girls (IQs of 70 to 90) after practice on a 45-rpm rotary pursuit task. Subjects were 96 girls from 7 to 17 yr. old. Each performed 14 trials on a rotary pursuit task (30-sec. trials, 10 sec. between trials), half performing the first seven trials with the nonpreferred hand, using the preferred hand on the next seven trials. The order was reversed for the remaining subjects. Nonsignificant differences between Trial 1 scores of the two groups indicated that the task was novel. Trial 1 scores of both groups were positively associated with age ( r = 0.5). There was no transfer to preferred hand, with negative transfer occurring to the nonpreferred hand. It was concluded that, for the task used in this study, mentally retarded girls do not experience positive bilateral transfer as do normal, age-matched girls.
25
Meaney, Karen S. "Developmental Modeling Effects on the Acquisition, Retention, and Transfer of a Novel Motor Task." Research Quarterly for Exercise and Sport 65, no. 1 (March 1994): 31–39. http://dx.doi.org/10.1080/02701367.1994.10762205.
Full text
26
Ramakrishnan, Ramya, Chongjie Zhang, and Julie Shah. "Perturbation Training for Human-Robot Teams." Journal of Artificial Intelligence Research 59 (July 31, 2017): 495–541. http://dx.doi.org/10.1613/jair.5390.
Full textAbstract:
In this work, we design and evaluate a computational learning model that enables a human-robot team to co-develop joint strategies for performing novel tasks that require coordination. The joint strategies are learned through "perturbation training," a human team-training strategy that requires team members to practice variations of a given task to help their team generalize to new variants of that task. We formally define the problem of human-robot perturbation training and develop and evaluate the first end-to-end framework for such training, which incorporates a multi-agent transfer learning algorithm, human-robot co-learning framework and communication protocol. Our transfer learning algorithm, Adaptive Perturbation Training (AdaPT), is a hybrid of transfer and reinforcement learning techniques that learns quickly and robustly for new task variants. We empirically validate the benefits of AdaPT through comparison to other hybrid reinforcement and transfer learning techniques aimed at transferring knowledge from multiple source tasks to a single target task. We also demonstrate that AdaPT's rapid learning supports live interaction between a person and a robot, during which the human-robot team trains to achieve a high level of performance for new task variants. We augment AdaPT with a co-learning framework and a computational bi-directional communication protocol so that the robot can co-train with a person during live interaction. Results from large-scale human subject experiments (n=48) indicate that AdaPT enables an agent to learn in a manner compatible with a human's own learning process, and that a robot undergoing perturbation training with a human results in a high level of team performance. Finally, we demonstrate that human-robot training using AdaPT in a simulation environment produces effective performance for a team incorporating an embodied robot partner.
27
Chen, Mengting, Yuxin Fang, Xinggang Wang, Heng Luo, Yifeng Geng, Xinyu Zhang, Chang Huang, Wenyu Liu, and Bo Wang. "Diversity Transfer Network for Few-Shot Learning." Proceedings of the AAAI Conference on Artificial Intelligence 34, no. 07 (April 3, 2020): 10559–66. http://dx.doi.org/10.1609/aaai.v34i07.6628.
Full textAbstract:
Few-shot learning is a challenging task that aims at training a classifier for unseen classes with only a few training examples. The main difficulty of few-shot learning lies in the lack of intra-class diversity within insufficient training samples. To alleviate this problem, we propose a novel generative framework, Diversity Transfer Network (DTN), that learns to transfer latent diversities from known categories and composite them with support features to generate diverse samples for novel categories in feature space. The learning problem of the sample generation (i.e., diversity transfer) is solved via minimizing an effective meta-classification loss in a single-stage network, instead of the generative loss in previous works. Besides, an organized auxiliary task co-training over known categories is proposed to stabilize the meta-training process of DTN. We perform extensive experiments and ablation studies on three datasets, i.e., miniImageNet, CIFAR100 and CUB. The results show that DTN, with single-stage training and faster convergence speed, obtains the state-of-the-art results among the feature generation based few-shot learning methods. Code and supplementary material are available at: https://github.com/Yuxin-CV/DTN.
28
Edwards, Jacqueline M., Digby Elliott, and Timothy D. Lee. "Contextual Interference Effects during Skill Acquisition and Transfer in Down’s Syndrome Adolescents." Adapted Physical Activity Quarterly 3, no. 3 (July 1986): 250–58. http://dx.doi.org/10.1123/apaq.3.3.250.
Full textAbstract:
An experiment is reported that investigated the effects of contextual interference on motor skill acquisition, and transfer of training in Down’s syndrome adolescents. Twenty Down’s syndrome adolescents and 20 nonhandicapped mental age controls learned a coincident anticipation timing task using either a random or a blocked training schedule. For transfer to a novel but similar task, subjects from both populations evidenced beneficial effects due to random practice. These data are discussed in terms of recent developments for strategy enhancement in motor learning by mentally retarded individuals.
29
Gulbaz, Rohail, Abdul Basit Siddiqui, Nadeem Anjum, Abdullah Alhumaidi Alotaibi, Turke Althobaiti, and Naeem Ramzan. "Balancer Genetic Algorithm—A Novel Task Scheduling Optimization Approach in Cloud Computing." Applied Sciences 11, no. 14 (July 6, 2021): 6244. http://dx.doi.org/10.3390/app11146244.
Full textAbstract:
Task scheduling is one of the core issues in cloud computing. Tasks are heterogeneous, and they have intensive computational requirements. Tasks need to be scheduled on Virtual Machines (VMs), which are resources in a cloud environment. Due to the immensity of search space for possible mappings of tasks to VMs, meta-heuristics are introduced for task scheduling. In scheduling makespan and load balancing, Quality of Service (QoS) parameters are crucial. This research contributes a novel load balancing scheduler, namely Balancer Genetic Algorithm (BGA), which is presented to improve makespan and load balancing. Insufficient load balancing can cause an overhead of utilization of resources, as some of the resources remain idle. BGA inculcates a load balancing mechanism, where the actual load in terms of million instructions assigned to VMs is considered. A need to opt for multi-objective optimization for improvement in load balancing and makespan is also emphasized. Skewed, normal and uniform distributions of workload and different batch sizes are used in experimentation. BGA has exhibited significant improvement compared with various state-of-the-art approaches for makespan, throughput and load balancing.
30
Xu, Yao, Xueshuang Xiang, and Meiyu Huang. "Task-Driven Common Representation Learning via Bridge Neural Network." Proceedings of the AAAI Conference on Artificial Intelligence 33 (July 17, 2019): 5573–80. http://dx.doi.org/10.1609/aaai.v33i01.33015573.
Full textAbstract:
This paper introduces a novel deep learning based method, named bridge neural network (BNN) to dig the potential relationship between two given data sources task by task. The proposed approach employs two convolutional neural networks that project the two data sources into a feature space to learn the desired common representation required by the specific task. The training objective with artificial negative samples is introduced with the ability of mini-batch training and it’s asymptotically equivalent to maximizing the total correlation of the two data sources, which is verified by the theoretical analysis. The experiments on the tasks, including pair matching, canonical correlation analysis, transfer learning, and reconstruction demonstrate the state-of-the-art performance of BNN, which may provide new insights into the aspect of common representation learning.
31
FOROODI-NEJAD, FARZANEH, and JOHANNE PARADIS. "Crosslinguistic transfer in the acquisition of compound words in Persian–English bilinguals." Bilingualism: Language and Cognition 12, no. 4 (September 16, 2009): 411–27. http://dx.doi.org/10.1017/s1366728909990241.
Full textAbstract:
Crosslinguistic transfer in bilingual language acquisition has been widely reported in various linguistic domains (e.g., Döpke, 1998; Nicoladis, 1999; Paradis, 2001). In this study we examined structural overlap (Döpke, 2000; Müller and Hulk, 2001) and dominance (Yip and Matthews, 2000) as explanatory factors for crosslinguistic transfer in Persian–English bilingual children's production of novel compound words. Nineteen Persian monolinguals, sixteen Persian–English bilinguals, and seventeen English monolinguals participated in a novel compound production task. Our results showed crosslinguistic influence of Persian on English and of English on Persian. Bilingual children produced more right-headed compounds in Persian, compared with Persian monolinguals, and in their English task, they produced more left-headed compounds than English monolinguals. Furthermore, Persian-dominant bilinguals tended more towards left-headed compounds in Persian than the English-dominant group. These findings point to both structural overlap and language dominance as factors underlying crosslinguistic transfer.
32
Ackerman, Phillip L. "Skill Acquisition, Individual Differences, and Human Abilities." Proceedings of the Human Factors Society Annual Meeting 30, no. 3 (September 1986): 270–74. http://dx.doi.org/10.1177/154193128603000316.
Full textAbstract:
The nature of individual differences in novel and practiced performance on skill acquisition tasks is considered from an information processing framework that incorporates concepts derived from automatic/controlled processing and attentional resource perspectives. A set of skill acquisition experiments graphically demonstrate changes in individual differences parameters via manipulating task characteristics of 1) information processing consistency, 2) memory load, 3) stimulus novelty. A further experiment illustrates the effects of novel, but consistent information processing demands on abilities, within a transfer-of-training paradigm. Results are discussed in the context of ability/skill relations.
33
Benavides-Prado, Diana. "An SVM-Based Framework for Long-Term Learning Systems." Proceedings of the AAAI Conference on Artificial Intelligence 33 (July 17, 2019): 9915–16. http://dx.doi.org/10.1609/aaai.v33i01.33019915.
Full textAbstract:
In our research, we study the problem of learning a sequence of supervised tasks. This is a long-standing challenge in machine learning. Our work relies on transfer of knowledge between hypotheses learned with Support Vector Machines. Transfer occurs in two directions: forward and backward. We have proposed to selectively transfer forward support vector coefficients from previous hypotheses as upper-bounds on support vector coefficients to be learned on a target task. We also proposed a novel method for refining existing hypotheses by transferring backward knowledge from a target hypothesis learned recently. We have improved this method through a hypothesis refinement approach that refines whilst encouraging retention of knowledge. Our contribution is represented in a long-term learning framework for binary classification tasks received sequentially one at a time.
34
Le, Duong, My Thai, and Thien Nguyen. "Multi-Task Learning for Metaphor Detection with Graph Convolutional Neural Networks and Word Sense Disambiguation." Proceedings of the AAAI Conference on Artificial Intelligence 34, no. 05 (April 3, 2020): 8139–46. http://dx.doi.org/10.1609/aaai.v34i05.6326.
Full textAbstract:
The current deep learning works on metaphor detection have only considered this task independently, ignoring the useful knowledge from the related tasks and knowledge resources. In this work, we introduce two novel mechanisms to improve the performance of the deep learning models for metaphor detection. The first mechanism employs graph convolutional neural networks (GCN) with dependency parse trees to directly connect the words of interest with their important context words for metaphor detection. The GCN networks in this work also present a novel control mechanism to filter the learned representation vectors to retain the most important information for metaphor detection. The second mechanism, on the other hand, features a multi-task learning framework that exploits the similarity between word sense disambiguation and metaphor detection to transfer the knowledge between the two tasks. The extensive experiments demonstrate the effectiveness of the proposed techniques, yielding the state-of-the-art performance over several datasets.
35
Li, Zheng, Ying Wei, Yu Zhang, Xiang Zhang, and Xin Li. "Exploiting Coarse-to-Fine Task Transfer for Aspect-Level Sentiment Classification." Proceedings of the AAAI Conference on Artificial Intelligence 33 (July 17, 2019): 4253–60. http://dx.doi.org/10.1609/aaai.v33i01.33014253.
Full textAbstract:
Aspect-level sentiment classification (ASC) aims at identifying sentiment polarities towards aspects in a sentence, where the aspect can behave as a general Aspect Category (AC) or a specific Aspect Term (AT). However, due to the especially expensive and labor-intensive labeling, existing public corpora in AT-level are all relatively small. Meanwhile, most of the previous methods rely on complicated structures with given scarce data, which largely limits the efficacy of the neural models. In this paper, we exploit a new direction named coarse-to-fine task transfer, which aims to leverage knowledge learned from a rich-resource source domain of the coarse-grained AC task, which is more easily accessible, to improve the learning in a low-resource target domain of the fine-grained AT task. To resolve both the aspect granularity inconsistency and feature mismatch between domains, we propose a Multi-Granularity Alignment Network (MGAN). In MGAN, a novel Coarse2Fine attention guided by an auxiliary task can help the AC task modeling at the same finegrained level with the AT task. To alleviate the feature false alignment, a contrastive feature alignment method is adopted to align aspect-specific feature representations semantically. In addition, a large-scale multi-domain dataset for the AC task is provided. Empirically, extensive experiments demonstrate the effectiveness of the MGAN.
36
Winne, Peter S. "Memory-Encoding Strategies and Concurrent-Task Practice: New Implications for Complex Skill Training." Proceedings of the Human Factors Society Annual Meeting 31, no. 6 (September 1987): 657–61. http://dx.doi.org/10.1177/154193128703100610.
Full textAbstract:
This study investigates memory-encoding strategies in a multiple-task environment. Eighty subjects solved mental arithmetic and trigram items in a transfer of training study. During training, practice load and variety were manipulated between groups. During transfer, the subjects solved rehearsed and novel items under single-, dual- and triple-task loads. Both main and interactive effects for practice load and variety were found. Variety influenced solution times for new and rehearsed items and these effects were moderated by practice load within levels of task load. The results are discussed within the framework of memory-encoding strategies, as applied to the design of training in complex systems.
37
Wang, Tony S. L., and Joo-Hyun Song. "Impaired visuomotor generalization by inconsistent attentional contexts." Journal of Neurophysiology 118, no. 3 (September 1, 2017): 1709–19. http://dx.doi.org/10.1152/jn.00089.2017.
Full textAbstract:
In daily life, people are constantly presented with situations in which they have to learn and acquire new motor skills in complex environments, where attention is often distracted by other events. Being able to generalize and perform the acquired motor action in different environments is a crucial part of visuomotor learning. The current study examined whether attentional distraction impairs generalization of visuomotor adaptation or whether consistent distraction can operate as an internal cue to facilitate generalization. Using a dual-task paradigm combining visuomotor rotational adaptation and an attention-demanding secondary task, we showed that switching the attentional context from training (dual-task) to generalization (single-task) reduced the range of transfer of visuomotor adaptation to untrained directions. However, when consistent distraction was present throughout training and generalization, visuomotor generalization was equivalent to without distractions at all. Furthermore, this attentional context-dependent generalization was evident even when sensory modality of distractions differed between training and generalization. Therefore, the general nature of the dual tasks, rather than the specific stimuli, is associated with visuomotor memory and serves as a critical cue for generalization. Taken together, we demonstrated that attention plays a critical role during sensorimotor adaptation in selecting and associating multisensory signals with motor memory. This finding provides insight into developing learning programs that are generalizable in complex daily environments. NEW & NOTEWORTHY Learning novel motor actions in complex environments with attentional distraction is a critical function. Successful motor learning involves the ability to transfer the acquired skill from the trained to novel environments. Here, we demonstrate attentional distraction does not impair visuomotor adaptation. Rather, consistency in the attentional context from training to generalization modulates the degree of transfer to untrained locations. The role of attention and memory must, therefore, be incorporated into existing models of visuomotor learning.
38
Talsma, Lotte J., Henryk A. Kroese, and Heleen A. Slagter. "Boosting Cognition: Effects of Multiple-Session Transcranial Direct Current Stimulation on Working Memory." Journal of Cognitive Neuroscience 29, no. 4 (April 2017): 755–68. http://dx.doi.org/10.1162/jocn_a_01077.
Full textAbstract:
Transcranial direct current stimulation (tDCS) is a promising tool for neurocognitive enhancement. Several studies have shown that just a single session of tDCS over the left dorsolateral pFC (lDLPFC) can improve the core cognitive function of working memory (WM) in healthy adults. Yet, recent studies combining multiple sessions of anodal tDCS over lDLPFC with verbal WM training did not observe additional benefits of tDCS in subsequent stimulation sessions nor transfer of benefits to novel WM tasks posttraining. Using an enhanced stimulation protocol as well as a design that included a baseline measure each day, the current study aimed to further investigate the effects of multiple sessions of tDCS on WM. Specifically, we investigated the effects of three subsequent days of stimulation with anodal (20 min, 1 mA) versus sham tDCS (1 min, 1 mA) over lDLPFC (with a right supraorbital reference) paired with a challenging verbal WM task. WM performance was measured with a verbal WM updating task (the letter n-back) in the stimulation sessions and several WM transfer tasks (different letter set n-back, spatial n-back, operation span) before and 2 days after stimulation. Anodal tDCS over lDLPFC enhanced WM performance in the first stimulation session, an effect that remained visible 24 hr later. However, no further gains of anodal tDCS were observed in the second and third stimulation sessions, nor did benefits transfer to other WM tasks at the group level. Yet, interestingly, post hoc individual difference analyses revealed that in the anodal stimulation group the extent of change in WM performance on the first day of stimulation predicted pre to post changes on both the verbal and the spatial transfer task. Notably, this relationship was not observed in the sham group. Performance of two individuals worsened during anodal stimulation and on the transfer tasks. Together, these findings suggest that repeated anodal tDCS over lDLPFC combined with a challenging WM task may be an effective method to enhance domain-independent WM functioning in some individuals, but not others, or can even impair WM. They thus call for a thorough investigation into individual differences in tDCS respondence as well as further research into the design of multisession tDCS protocols that may be optimal for boosting cognition across a wide range of individuals.
39
Benavides-Prado, Diana, Yun Sing Koh, and Patricia Riddle. "Towards Knowledgeable Supervised Lifelong Learning Systems." Journal of Artificial Intelligence Research 68 (May 8, 2020): 159–224. http://dx.doi.org/10.1613/jair.1.11432.
Full textAbstract:
Learning a sequence of tasks is a long-standing challenge in machine learning. This setting applies to learning systems that observe examples of a range of tasks at different points in time. A learning system should become more knowledgeable as more related tasks are learned. Although the problem of learning sequentially was acknowledged for the first time decades ago, the research in this area has been rather limited. Research in transfer learning, multitask learning, metalearning and deep learning has studied some challenges of these kinds of systems. Recent research in lifelong machine learning and continual learning has revived interest in this problem. We propose Proficiente, a full framework for long-term learning systems. Proficiente relies on knowledge transferred between hypotheses learned with Support Vector Machines. The first component of the framework is focused on transferring forward selectively from a set of existing hypotheses or functions representing knowledge acquired during previous tasks to a new target task. A second component of Proficiente is focused on transferring backward, a novel ability of long-term learning systems that aim to exploit knowledge derived from recent tasks to encourage refinement of existing knowledge. We propose a method that transfers selectively from a task learned recently to existing hypotheses representing previous tasks. The method encourages retention of existing knowledge whilst refining. We analyse the theoretical properties of the proposed framework. Proficiente is accompanied by an agnostic metric that can be used to determine if a long-term learning system is becoming more knowledgeable. We evaluate Proficiente in both synthetic and real-world datasets, and demonstrate scenarios where knowledgeable supervised learning systems can be achieved by means of transfer.
40
Fraulini, Nicholas W., Monica A. Perez, Teresa L. Perez, Alexa L. Fistel, and James L. Szalma. "A Preliminary Study Examining Novel Training Paradigms for Vigilance." Proceedings of the Human Factors and Ergonomics Society Annual Meeting 61, no. 1 (September 2017): 1509–13. http://dx.doi.org/10.1177/1541931213601862.
Full textAbstract:
Training for vigilance has been a primary research question for over 70 years. Specifically, researchers have fought to lessen the effects of the vigilance decrement, or the typical decline in performance as time on task increases. In the present study, we examine two forms of training for vigilance: practice and knowledge of result (KR). We propose that providing observers with either practice, KR, or a combination of the two during training will improve performance on a transfer vigil. Our results showed observers receiving practice displayed higher sensitivity and increased conservatism during training, as well as a trend toward higher sensitivity during transfer. These results show the benefits of providing observers practice, which include their performance on a transfer vigil as well as the efficiency of the training itself. We discuss the implications of these findings and how they may impact training for vigilance in the future.
41
Dunne, Stephen, Amanda Ellison, and Daniel T. Smith. "The Limitations of Reward Effects on Saccade Latencies: An Exploration of Task-Specificity and Strength." Vision 3, no. 2 (May 11, 2019): 20. http://dx.doi.org/10.3390/vision3020020.
Full textAbstract:
Saccadic eye movements are simple, visually guided actions. Operant conditioning of specific saccade directions can reduce the latency of eye movements in the conditioned direction. However, it is not clear to what extent this learning transfers from the conditioned task to novel tasks. The purpose of this study was to investigate whether the effects of operant conditioning of prosaccades to specific spatial locations would transfer to more complex oculomotor behaviours, specifically, prosaccades made in the presence of a distractor (Experiment 1) and antisaccades (Experiment 2). In part 1 of each experiment, participants were rewarded for making a saccade to one hemifield. In both experiments, the reward produced a significant facilitation of saccadic latency for prosaccades directed to the rewarded hemifield. In part 2, rewards were withdrawn, and the participant made a prosaccade to targets that were accompanied by a contralateral distractor (Experiment 1) or an antisaccade (Experiment 2). There were no hemifield-specific effects of the reward on saccade latency on the remote distractor effect or antisaccades, although the reward was associated with an overall slowing of saccade latency in Experiment 1. These data indicate that operant conditioning of saccadic eye movements does not transfer to similar but untrained tasks. We conclude that rewarding specific spatial locations is unlikely to induce long-term, systemic changes to the human oculomotor system.
42
Leber, Leray L., Christopher D. Wickens, Christopher Bakke, Michael Sulek, and William Marshak. "Voice and Manual Control in Dual Task Situations." Proceedings of the Human Factors Society Annual Meeting 31, no. 4 (September 1987): 419–23. http://dx.doi.org/10.1177/154193128703100408.
Full textAbstract:
The objective of this research was to replicate and extend an investigation of Voice and Manual Control in Dual Task Situations previously performed by Wickens et al. (1985). This study incorporated both the previous within-subject design with a much larger sample size and a novel between-subject paradigm. The repeated measures investigation minimizing asymmetric transfer between response conditions revealed significantly better performance when a verbal Sternberg task was voice controlled in combination with a manually controlled spatial tracking task. The between-subject study likewise supported this finding. The previous 1985 study's findings favoring hemispherically compatible left-handed tracking were not supported in this investigation.
43
Choi, Seung-oh, Harry J. Meeuwsen, Ron French, Claudine Sherrill, and Rozie McCabe. "Motor Skill Acquisition, Retention, and Transfer in Adults with Profound Mental Retardation." Adapted Physical Activity Quarterly 18, no. 3 (July 2001): 257–72. http://dx.doi.org/10.1123/apaq.18.3.257.
Full textAbstract:
The purpose was to examine whether adults with profound mental retardation (PMR) have the ability to learn and transfer a motor skill to a novel situation. In Experiment 1, novel task transfer performance was examined. Six male adults with PMR threw beanbags three different distances during acquisition, followed by four novel transfer distances and a novel implement (a horse shoe). In Experiment 2, a 48-hr and a 1-week delayed retention test was used with 6 different males with PMR who practiced three beanbag-throwing distances and then performed two familiar and two novel distances for each retention test. Analyses indicated that, with concurrent visual information of the target, adults with PMR can throw accurately on retention and transfer tests and can generalize beanbag throwing skill to horseshoe-throwing. The prototype model of memory representation seems to explain the findings better than the exemplar model. In addition, random practice of skill variations appears to be an effective teaching strategy.
44
Liaw, Rung-Tzuo, and Chuan-Kang Ting. "Evolutionary Manytasking Optimization Based on Symbiosis in Biocoenosis." Proceedings of the AAAI Conference on Artificial Intelligence 33 (July 17, 2019): 4295–303. http://dx.doi.org/10.1609/aaai.v33i01.33014295.
Full textAbstract:
Evolutionary multitasking is a significant emerging search paradigm that utilizes evolutionary algorithms to concurrently optimize multiple tasks. The multi-factorial evolutionary algorithm renders an effectual realization of evolutionary multitasking on two or three tasks. However, there remains room for improvement on the performance and capability of evolutionary multitasking. Beyond three tasks, this paper proposes a novel framework, called the symbiosis in biocoenosis optimization (SBO), to address evolutionary many-tasking optimization. The SBO leverages the notion of symbiosis in biocoenosis for transferring information and knowledge among different tasks through three major components: 1) transferring information through inter-task individual replacement, 2) measuring symbiosis through intertask paired evaluations, and 3) coordinating the frequency and quantity of transfer based on symbiosis in biocoenosis. The inter-task individual replacement with paired evaluations caters for estimation of symbiosis, while the symbiosis in biocoenosis provides a good estimator of transfer. This study examines the effectiveness and efficiency of the SBO on a suite of many-tasking benchmark problems, designed to deal with 30 tasks simultaneously. The experimental results show that SBO leads to better solutions and faster convergence than the state-of-the-art evolutionary multitasking algorithms. Moreover, the results indicate that SBO is highly capable of identifying the similarity between problems and transferring information appropriately.
45
Cooper, Joel. "Do Driving Impairments from Concurrent Cell-Phone Use Diminish with Practice?" Proceedings of the Human Factors and Ergonomics Society Annual Meeting 51, no. 24 (October 2007): 1536–39. http://dx.doi.org/10.1177/154193120705102405.
Full textAbstract:
Our research examined the effects of practice on in-vehicle cell-phone use. Drivers that reported either high or low real world cell-phone usage were selected to participate in four, hour-and-a-half simulated driving sessions, on different days. The research consisted of two phases, a training phase and a novel transfer phase. Compared to single-task driving, dual-task performance deficits persisted through training and transfer driving conditions. Furthermore, groups of high and low real world experience were equally impaired. It is concluded that practice does not improve the ability to drive while conversing on a cellphone.
46
Poh, Eugene, Timothy J. Carroll, and Jordan A. Taylor. "Effect of coordinate frame compatibility on the transfer of implicit and explicit learning across limbs." Journal of Neurophysiology 116, no. 3 (September 1, 2016): 1239–49. http://dx.doi.org/10.1152/jn.00410.2016.
Full textAbstract:
Insights into the neural representation of motor learning can be obtained by investigating how learning transfers to novel task conditions. We recently demonstrated that visuomotor rotation learning transferred strongly between left and right limbs when the task was performed in a sagittal workspace, which afforded a consistent remapping for the two limbs in both extrinsic and joint-based coordinates. In contrast, transfer was absent when performed in horizontal workspace, where the extrinsically defined perturbation required conflicting joint-based remapping for the left and right limbs. Because visuomotor learning is thought to be supported by both implicit and explicit forms of learning, however, it is unclear to what extent these distinct forms of learning contribute to interlimb transfer. In this study, we assessed the degree to which interlimb transfer, following visuomotor rotation training, reflects explicit vs. implicit learning by obtaining verbal reports of participants' aiming direction before each movement. We also determined the extent to which these distinct components of learning are constrained by the compatibility of coordinate systems by comparing transfer between groups of participants who reached to targets arranged in the horizontal and sagittal planes. Both sagittal and horizontal conditions displayed complete transfer of explicit learning to the untrained limb. In contrast, transfer of implicit learning was incomplete, but the sagittal condition showed greater transfer than the horizontal condition. These findings suggest that explicit strategies developed with one limb can be fully implemented in the opposite limb, whereas implicit transfer depends on the degree to which new sensorimotor maps are spatially compatible for the two limbs.
47
Chen, Riquan, Tianshui Chen, Xiaolu Hui, Hefeng Wu, Guanbin Li, and Liang Lin. "Knowledge Graph Transfer Network for Few-Shot Recognition." Proceedings of the AAAI Conference on Artificial Intelligence 34, no. 07 (April 3, 2020): 10575–82. http://dx.doi.org/10.1609/aaai.v34i07.6630.
Full textAbstract:
Few-shot learning aims to learn novel categories from very few samples given some base categories with sufficient training samples. The main challenge of this task is the novel categories are prone to dominated by color, texture, shape of the object or background context (namely specificity), which are distinct for the given few training samples but not common for the corresponding categories (see Figure 1). Fortunately, we find that transferring information of the correlated based categories can help learn the novel concepts and thus avoid the novel concept being dominated by the specificity. Besides, incorporating semantic correlations among different categories can effectively regularize this information transfer. In this work, we represent the semantic correlations in the form of structured knowledge graph and integrate this graph into deep neural networks to promote few-shot learning by a novel Knowledge Graph Transfer Network (KGTN). Specifically, by initializing each node with the classifier weight of the corresponding category, a propagation mechanism is learned to adaptively propagate node message through the graph to explore node interaction and transfer classifier information of the base categories to those of the novel ones. Extensive experiments on the ImageNet dataset show significant performance improvement compared with current leading competitors. Furthermore, we construct an ImageNet-6K dataset that covers larger scale categories, i.e, 6,000 categories, and experiments on this dataset further demonstrate the effectiveness of our proposed model.
48
Lefumat, Hannah Z., Jean-Louis Vercher, R. Chris Miall, Jonathan Cole, Frank Buloup, Lionel Bringoux, Christophe Bourdin, and Fabrice R. Sarlegna. "To transfer or not to transfer? Kinematics and laterality quotient predict interlimb transfer of motor learning." Journal of Neurophysiology 114, no. 5 (November 2015): 2764–74. http://dx.doi.org/10.1152/jn.00749.2015.
Full textAbstract:
Humans can remarkably adapt their motor behavior to novel environmental conditions, yet it remains unclear which factors enable us to transfer what we have learned with one limb to the other. Here we tested the hypothesis that interlimb transfer of sensorimotor adaptation is determined by environmental conditions but also by individual characteristics. We specifically examined the adaptation of unconstrained reaching movements to a novel Coriolis, velocity-dependent force field. Right-handed subjects sat at the center of a rotating platform and performed forward reaching movements with the upper limb toward flashed visual targets in prerotation, per-rotation (i.e., adaptation), and postrotation tests. Here only the dominant arm was used during adaptation and interlimb transfer was assessed by comparing performance of the nondominant arm before and after dominant-arm adaptation. Vision and no-vision conditions did not significantly influence interlimb transfer of trajectory adaptation, which on average was significant but limited. We uncovered a substantial heterogeneity of interlimb transfer across subjects and found that interlimb transfer can be qualitatively and quantitatively predicted for each healthy young individual. A classifier showed that in our study, interlimb transfer could be predicted based on the subject's task performance, most notably motor variability during learning, and his or her laterality quotient. Positive correlations suggested that variability of motor performance and lateralization of arm movement control facilitate interlimb transfer. We further show that these individual characteristics can predict the presence and the magnitude of interlimb transfer of left-handers. Overall, this study suggests that individual characteristics shape the way the nervous system can generalize motor learning.
49
König, Matthias, Gaspar Epro, John Seeley, Wolfgang Potthast, and Kiros Karamanidis. "Retention and generalizability of balance recovery response adaptations from trip perturbations across the adult life span." Journal of Neurophysiology 122, no. 5 (November 1, 2019): 1884–93. http://dx.doi.org/10.1152/jn.00380.2019.
Full textAbstract:
For human locomotion, varying environments require adjustments of the motor system. We asked whether age affects gait balance recovery adaptation, its retention over months, and the transfer of adaptation to an untrained reactive balance task. Healthy adults (26 young, 27 middle-aged, and 25 older; average ages 24, 52, and 72 yr, respectively) completed two tasks. The primary task involved treadmill walking: either unperturbed (control; n = 39) or subject to unexpected trip perturbations (training; n = 39). A single trip perturbation was repeated after a 14-wk retention period. The secondary transfer task, before and after treadmill walking, involved sudden loss of balance in a lean-and-release protocol. For both tasks, the anteroposterior margin of stability (MoS) was calculated at foot touchdown. For the first (i.e., novel) trip, older adults required one more recovery step ( P = 0.03) to regain positive MoS compared with younger, but not middle-aged, adults. However, over several trip perturbations, all age groups increased their MoS for the first recovery step to a similar extent (up to 70%) and retained improvements over 14 wk, although a decay over time was found for older adults ( P = 0.002; middle-aged showing a tendency for decay: P = 0.076). Thus, although adaptability in reactive gait stability control remains effective across the adult life span, retention of adaptations over time appears diminished with aging. Despite these robust adaptations, the perturbation training group did not show superior improvements in the transfer task compared with age-matched controls (no differences in MoS changes), suggesting that generalizability of acquired fall-resisting skills from gait-perturbation training may be limited. NEW & NOTEWORTHY The human neuromotor system preserves its adaptability across the adult life span. However, although adaptability in reactive gait stability control remains effective as age increases, retention of recovery response adaptations over time appears to be reduced with aging. Furthermore, acquired fall-resisting skills from single-session perturbation training seem task specific, which may limit the generalizability of such training to the variety of real-life falls.
50
Tanaka, James W., Tim Curran, and David L. Sheinberg. "The Training and Transfer of Real-World Perceptual Expertise." Psychological Science 16, no. 2 (February 2005): 145–51. http://dx.doi.org/10.1111/j.0956-7976.2005.00795.x.
Full textAbstract:
A hallmark of perceptual expertise is that experts classify objects at a more specific, subordinate level of abstraction than novices. To what extent does subordinate-level learning contribute to the transfer of perceptual expertise to novel exemplars and novel categories? In this study, participants learned to classify 10 varieties of wading birds and 10 varieties of owls at either the subordinate, species (e.g., “great blue crown heron,” “eastern screech owl”) or the family (“wading bird,” “owl”) level of abstraction. During training, the amount of visual exposure was equated such that participants received an equal number of learning trials for wading birds and owls. Pre- and posttraining performance was measured in a same/different discrimination task in which participants judged whether pairs of bird stimuli belonged to the same or different species. Participants trained in species-level discrimination demonstrated greater transfer to novel exemplars and novel species categories than participants trained in family-level discrimination. These findings suggest that perceptual categorization, not perceptual exposure per se, is important for the development and generalization of visual expertise.