Добірка наукової літератури з теми "DNN architecture"
Оформте джерело за APA, MLA, Chicago, Harvard та іншими стилями
Ознайомтеся зі списками актуальних статей, книг, дисертацій, тез та інших наукових джерел на тему "DNN architecture".
Біля кожної праці в переліку літератури доступна кнопка «Додати до бібліографії». Скористайтеся нею – і ми автоматично оформимо бібліографічне посилання на обрану працю в потрібному вам стилі цитування: APA, MLA, «Гарвард», «Чикаго», «Ванкувер» тощо.
Також ви можете завантажити повний текст наукової публікації у форматі «.pdf» та прочитати онлайн анотацію до роботи, якщо відповідні параметри наявні в метаданих.
Статті в журналах з теми "DNN architecture"
1
Roorda, Esther, Seyedramin Rasoulinezhad, Philip H. W. Leong, and Steven J. E. Wilton. "FPGA Architecture Exploration for DNN Acceleration." ACM Transactions on Reconfigurable Technology and Systems 15, no. 3 (September 30, 2022): 1–37. http://dx.doi.org/10.1145/3503465.
Повний текст джерелаАнотація:
Recent years have seen an explosion of machine learning applications implemented on Field-Programmable Gate Arrays (FPGAs) . FPGA vendors and researchers have responded by updating their fabrics to more efficiently implement machine learning accelerators, including innovations such as enhanced Digital Signal Processing (DSP) blocks and hardened systolic arrays. Evaluating architectural proposals is difficult, however, due to the lack of publicly available benchmark circuits. This paper addresses this problem by presenting an open-source benchmark circuit generator that creates realistic DNN-oriented circuits for use in FPGA architecture studies. Unlike previous generators, which create circuits that are agnostic of the underlying FPGA, our circuits explicitly instantiate embedded blocks, allowing for meaningful comparison of recent architectural proposals without the need for a complete inference computer-aided design (CAD) flow. Our circuits are compatible with the VTR CAD suite, allowing for architecture studies that investigate routing congestion and other low-level architectural implications. In addition to addressing the lack of machine learning benchmark circuits, the architecture exploration flow that we propose allows for a more comprehensive evaluation of FPGA architectures than traditional static benchmark suites. We demonstrate this through three case studies which illustrate how realistic benchmark circuits can be generated to target different heterogeneous FPGAs.
2
Elola, Andoni, Elisabete Aramendi, Unai Irusta, Artzai Picón, Erik Alonso, Pamela Owens, and Ahamed Idris. "Deep Neural Networks for ECG-Based Pulse Detection during Out-of-Hospital Cardiac Arrest." Entropy 21, no. 3 (March 21, 2019): 305. http://dx.doi.org/10.3390/e21030305.
Повний текст джерелаАнотація:
The automatic detection of pulse during out-of-hospital cardiac arrest (OHCA) is necessary for the early recognition of the arrest and the detection of return of spontaneous circulation (end of the arrest). The only signal available in every single defibrillator and valid for the detection of pulse is the electrocardiogram (ECG). In this study we propose two deep neural network (DNN) architectures to detect pulse using short ECG segments (5 s), i.e., to classify the rhythm into pulseless electrical activity (PEA) or pulse-generating rhythm (PR). A total of 3914 5-s ECG segments, 2372 PR and 1542 PEA, were extracted from 279 OHCA episodes. Data were partitioned patient-wise into training (80%) and test (20%) sets. The first DNN architecture was a fully convolutional neural network, and the second architecture added a recurrent layer to learn temporal dependencies. Both DNN architectures were tuned using Bayesian optimization, and the results for the test set were compared to state-of-the art PR/PEA discrimination algorithms based on machine learning and hand crafted features. The PR/PEA classifiers were evaluated in terms of sensitivity (Se) for PR, specificity (Sp) for PEA, and the balanced accuracy (BAC), the average of Se and Sp. The Se/Sp/BAC of the DNN architectures were 94.1%/92.9%/93.5% for the first one, and 95.5%/91.6%/93.5% for the second one. Both architectures improved the performance of state of the art methods by more than 1.5 points in BAC.
3
Tran, Van Duy, Duc Khai Lam, and Thi Hong Tran. "Hardware-Based Architecture for DNN Wireless Communication Models." Sensors 23, no. 3 (January 23, 2023): 1302. http://dx.doi.org/10.3390/s23031302.
Повний текст джерелаАнотація:
Multiple Input Multiple Output Orthogonal Frequency Division Multiplexing (MIMO OFDM) is a key technology for wireless communication systems. However, because of the problem of a high peak-to-average power ratio (PAPR), OFDM symbols can be distorted at the MIMO OFDM transmitter. It degrades the signal detection and channel estimation performance at the MIMO OFDM receiver. In this paper, three deep neural network (DNN) models are proposed to solve the problem of non-linear distortions introduced by the power amplifier (PA) of the transmitters and replace the conventional digital signal processing (DSP) modules at the receivers in 2 × 2 MIMO OFDM and 4 × 4 MIMO OFDM systems. Proposed model type I uses the DNN model to de-map the signals at the receiver. Proposed model type II uses the DNN model to learn and filter out the channel noises at the receiver. Proposed model type III uses the DNN model to de-map and detect the signals at the receiver. All three model types attempt to solve the non-linear problem. The robust bit error rate (BER) performances of the proposed receivers are achieved through the software and hardware implementation results. In addition, we have also implemented appropriate hardware architectures for the proposed DNN models using special techniques, such as quantization and pipeline to check the feasibility in practice, which recent studies have not done. Our hardware architectures are successfully designed and implemented on the Virtex 7 vc709 FPGA board.
4
Turner, Daniel, Pedro J. S. Cardoso, and João M. F. Rodrigues. "Modular Dynamic Neural Network: A Continual Learning Architecture." Applied Sciences 11, no. 24 (December 18, 2021): 12078. http://dx.doi.org/10.3390/app112412078.
Повний текст джерелаАнотація:
Learning to recognize a new object after having learned to recognize other objects may be a simple task for a human, but not for machines. The present go-to approaches for teaching a machine to recognize a set of objects are based on the use of deep neural networks (DNN). So, intuitively, the solution for teaching new objects on the fly to a machine should be DNN. The problem is that the trained DNN weights used to classify the initial set of objects are extremely fragile, meaning that any change to those weights can severely damage the capacity to perform the initial recognitions; this phenomenon is known as catastrophic forgetting (CF). This paper presents a new (DNN) continual learning (CL) architecture that can deal with CF, the modular dynamic neural network (MDNN). The presented architecture consists of two main components: (a) the ResNet50-based feature extraction component as the backbone; and (b) the modular dynamic classification component, which consists of multiple sub-networks and progressively builds itself up in a tree-like structure that rearranges itself as it learns over time in such a way that each sub-network can function independently. The main contribution of the paper is a new architecture that is strongly based on its modular dynamic training feature. This modular structure allows for new classes to be added while only altering specific sub-networks in such a way that previously known classes are not forgotten. Tests on the CORe50 dataset showed results above the state of the art for CL architectures.
5
Lee, Junghwan, Huanli Sun, Yuxia Liu, Xue Li, Yixin Liu, and Myungjun Kim. "State-of-Health Estimation and Anomaly Detection in Li-Ion Batteries Based on a Novel Architecture with Machine Learning." Batteries 9, no. 5 (May 8, 2023): 264. http://dx.doi.org/10.3390/batteries9050264.
Повний текст джерелаАнотація:
Variations across cells, modules, packs, and vehicles can cause significant errors in the state estimation of LIBs using machine learning algorithms, especially when trained with small datasets. Training with large datasets that account for all variations is often impractical due to resource and time constraints at initial product release. To address this issue, we proposed a novel architecture that leverages electronic control units, edge computers, and the cloud to detect unrevealed variations and abnormal degradations in LIBs. The architecture comprised a generalized deep neural network (DNN) for generalizability, a personalized DNN for accuracy within a vehicle, and a detector. We emphasized that a generalized DNN trained with small datasets must show reasonable estimation accuracy during cross validation, which is critical for real applications before online training. We demonstrated the feasibility of the architecture by conducting experiments on 65 DNN models, where we found distinct hyperparameter configurations. The results showed that the personalized DNN achieves a root mean square error (RMSE) of 0.33%, while the generalized DNN achieves an RMSE of 4.6%. Finally, the Mahalanobis distance was used to consider the SOH differences between the generalized DNN and personalized DNN to detect abnormal degradations.
6
Mudgil, Pooja, Pooja Gupta, Iti Mathur, and Nisheeth Joshi. "An ontological architecture for context data retrieval and ranking using SVM and DNN." Journal of Information and Optimization Sciences 44, no. 3 (2023): 369–82. http://dx.doi.org/10.47974/jios-1347.
Повний текст джерелаАнотація:
Context retrieval and ranking have always been an area of interest for researchers around the world. The ranking provides significance to the data that has to be presented in front of users but it also consumes time if the ranking architecture is not organized. The retrieval is dependent upon the co-relation among the data attributes that are supplied against a class label also referred to as ground truth and the ranking depends upon the sensing polarity that indicates the hold of the outcome towards asked information. This paper illustrates an ontological architecture that involves two phases namely context retrieval and ranking. The ranking phase is composed of three different algorithm architectures namely k-means, Support Vector Machines (SVM), and Deep Neural Networks (DNN). The DNN is tuned to fit and work as per the availability of a total number of samples. The proposed work has been evaluated for both quantitative and qualitative parameters in different sets and scenarios. The proposed work has also been compared with other state of art techniques and is illustrated in the paper itself.
7
Elsisi, Mahmoud, and Minh-Quang Tran. "Development of an IoT Architecture Based on a Deep Neural Network against Cyber Attacks for Automated Guided Vehicles." Sensors 21, no. 24 (December 18, 2021): 8467. http://dx.doi.org/10.3390/s21248467.
Повний текст джерелаАнотація:
This paper introduces an integrated IoT architecture to handle the problem of cyber attacks based on a developed deep neural network (DNN) with a rectified linear unit in order to provide reliable and secure online monitoring for automated guided vehicles (AGVs). The developed IoT architecture based on a DNN introduces a new approach for the online monitoring of AGVs against cyber attacks with a cheap and easy implementation instead of the traditional cyber attack detection schemes in the literature. The proposed DNN is trained based on experimental AGV data that represent the real state of the AGV and different types of cyber attacks including a random attack, ramp attack, pulse attack, and sinusoidal attack that is injected by the attacker into the internet network. The proposed DNN is compared with different deep learning and machine learning algorithms such as a one dimension convolutional neural network (1D-CNN), a supported vector machine model (SVM), random forest, extreme gradient boosting (XGBoost), and a decision tree for greater validation. Furthermore, the proposed IoT architecture based on a DNN can provide an effective detection for the AGV status with an excellent accuracy of 96.77% that is significantly greater than the accuracy based on the traditional schemes. The AGV status based on the proposed IoT architecture with a DNN is visualized by an advanced IoT platform named CONTACT Elements for IoT. Different test scenarios with a practical setup of an AGV with IoT are carried out to emphasize the performance of the suggested IoT architecture based on a DNN. The results approve the usefulness of the proposed IoT to provide effective cybersecurity for data visualization and tracking of the AGV status that enhances decision-making and improves industrial productivity.
8
P, Shanmugavadivu, Mary Shanthi Rani M, Chitra P, Lakshmanan S, Nagaraja P, and Vignesh U. "Bio-Optimization of Deep Learning Network Architectures." Security and Communication Networks 2022 (September 20, 2022): 1–11. http://dx.doi.org/10.1155/2022/3718340.
Повний текст джерелаАнотація:
Deep learning is reaching new heights as a result of its cutting-edge performance in a variety of fields, including computer vision, natural language processing, time series analysis, and healthcare. Deep learning is implemented using batch and stochastic gradient descent methods, as well as a few optimizers; however, this led to subpar model performance. However, there is now a lot of effort being done to improve deep learning’s performance using gradient optimization methods. The suggested work analyses convolutional neural networks (CNN) and deep neural networks (DNN) using several cutting-edge optimizers to enhance the performance of architectures. This work uses specific optimizers (SGD, RMSprop, Adam, Adadelta, etc.) to enhance the performance of designs using different types of datasets for result matching. A thorough report on the optimizers’ performance across a variety of architectures and datasets finishes the study effort. This research will be helpful to researchers in developing their framework and appropriate architecture optimizers. The proposed work involves eight new optimizers using four CNN and DNN architectures. The experimental results exploit breakthrough results for improving the efficiency of CNN and DNN architectures using various datasets.
9
Krishnan, Gokul, Sumit K. Mandal, Chaitali Chakrabarti, Jae-Sun Seo, Umit Y. Ogras, and Yu Cao. "Impact of On-chip Interconnect on In-memory Acceleration of Deep Neural Networks." ACM Journal on Emerging Technologies in Computing Systems 18, no. 2 (April 30, 2022): 1–22. http://dx.doi.org/10.1145/3460233.
Повний текст джерелаАнотація:
With the widespread use of Deep Neural Networks (DNNs), machine learning algorithms have evolved in two diverse directions—one with ever-increasing connection density for better accuracy and the other with more compact sizing for energy efficiency. The increase in connection density increases on-chip data movement, which makes efficient on-chip communication a critical function of the DNN accelerator. The contribution of this work is threefold. First, we illustrate that the point-to-point (P2P)-based interconnect is incapable of handling a high volume of on-chip data movement for DNNs. Second, we evaluate P2P and network-on-chip (NoC) interconnect (with a regular topology such as a mesh) for SRAM- and ReRAM-based in-memory computing (IMC) architectures for a range of DNNs. This analysis shows the necessity for the optimal interconnect choice for an IMC DNN accelerator. Finally, we perform an experimental evaluation for different DNNs to empirically obtain the performance of the IMC architecture with both NoC-tree and NoC-mesh. We conclude that, at the tile level, NoC-tree is appropriate for compact DNNs employed at the edge, and NoC-mesh is necessary to accelerate DNNs with high connection density. Furthermore, we propose a technique to determine the optimal choice of interconnect for any given DNN. In this technique, we use analytical models of NoC to evaluate end-to-end communication latency of any given DNN. We demonstrate that the interconnect optimization in the IMC architecture results in up to 6 × improvement in energy-delay-area product for VGG-19 inference compared to the state-of-the-art ReRAM-based IMC architectures.
10
Zhao, Jiaqi, Ming Xu, Yunzhi Chen, and Guoliang Xu. "A DNN Architecture Generation Method for DDoS Detection via Genetic Alogrithm." Future Internet 15, no. 4 (March 26, 2023): 122. http://dx.doi.org/10.3390/fi15040122.
Повний текст джерелаАнотація:
Nowdays, DNNs (Deep Neural Networks) are widely used in the field of DDoS attack detection. However, designing a good DNN architecture relies on the designer’s experience and requires considerable work. In this paper, a GA (genetic algorithm) is used to automatically generate the DNN architecture for DDoS detection to minimize human intervention in the design process. Furthermore, given the complexity of contemporary networks and the diversity of DDoS attacks, the objective of this paper is to generate a DNN model that boasts superior performance, real-time capability, and generalization ability to tackle intricate network scenarios. This paper presents a fitness function that guarantees the best model generated possesses a specific level of real-time capability. Additionally, the proposed method employs multiple datasets to joint models generated, thereby enhancing the model’s generalization performance. This paper conducts several experiments to validate the viability of the proposed method. Firstly, the best model generated with one dataset is compared with existing DNN models on the CICDDoS2019 dataset. The experimental results indicate that the model generated with one dataset has higher precision and F1-score than the existing DNN models. Secondly, model generation experiments are conducted on the CICIDS2017 and CICIDS2018 datasets, and the best model generated still performs well. Finally, this paper conducts comparative experiments on multiple datasets using the best model generated with six datasets and the best model generated by existing methods. The experimental results demonstrate that the best model generated with six datasets has better generalization ability and real-time capability.
Дисертації з теми "DNN architecture"
1
Azam, Md Ali. "Energy Efficient Spintronic Device for Neuromorphic Computation." VCU Scholars Compass, 2019. https://scholarscompass.vcu.edu/etd/6036.
Повний текст джерелаАнотація:
Future computing will require significant development in new computing device paradigms. This is motivated by CMOS devices reaching their technological limits, the need for non-Von Neumann architectures as well as the energy constraints of wearable technologies and embedded processors. The first device proposal, an energy-efficient voltage-controlled domain wall device for implementing an artificial neuron and synapse is analyzed using micromagnetic modeling. By controlling the domain wall motion utilizing spin transfer or spin orbit torques in association with voltage generated strain control of perpendicular magnetic anisotropy in the presence of Dzyaloshinskii-Moriya interaction (DMI), different positions of the domain wall are realized in the free layer of a magnetic tunnel junction to program different synaptic weights. Additionally, an artificial neuron can be realized by combining this DW device with a CMOS buffer. The second neuromorphic device proposal is inspired by the brain. Membrane potential of many neurons oscillate in a subthreshold damped fashion and fire when excited by an input frequency that nearly equals their Eigen frequency. We investigate theoretical implementation of such “resonate-and-fire” neurons by utilizing the magnetization dynamics of a fixed magnetic skyrmion based free layer of a magnetic tunnel junction (MTJ). Voltage control of magnetic anisotropy or voltage generated strain results in expansion and shrinking of a skyrmion core that mimics the subthreshold oscillation. Finally, we show that such resonate and fire neurons have potential application in coupled nanomagnetic oscillator based associative memory arrays.
2
Riera, Villanueva Marc. "Low-power accelerators for cognitive computing." Doctoral thesis, Universitat Politècnica de Catalunya, 2020. http://hdl.handle.net/10803/669828.
Повний текст джерелаАнотація:
Deep Neural Networks (DNNs) have achieved tremendous success for cognitive applications, and are especially efficient in classification and decision making problems such as speech recognition or machine translation. Mobile and embedded devices increasingly rely on DNNs to understand the world. Smartphones, smartwatches and cars perform discriminative tasks, such as face or object recognition, on a daily basis. Despite the increasing popularity of DNNs, running them on mobile and embedded systems comes with several main challenges: delivering high accuracy and performance with a small memory and energy budget. Modern DNN models consist of billions of parameters requiring huge computational and memory resources and, hence, they cannot be directly deployed on low-power systems with limited resources. The objective of this thesis is to address these issues and propose novel solutions in order to design highly efficient custom accelerators for DNN-based cognitive computing systems.
In first place, we focus on optimizing the inference of DNNs for sequence processing applications. We perform an analysis of the input similarity between consecutive DNN executions. Then, based on the high degree of input similarity, we propose DISC, a hardware accelerator implementing a Differential Input Similarity Computation technique to reuse the computations of the previous execution, instead of computing the entire DNN. We observe that, on average, more than 60% of the inputs of any neural network layer tested exhibit negligible changes with respect to the previous execution. Avoiding the memory accesses and computations for these inputs results in 63% energy savings on average.
In second place, we propose to further optimize the inference of FC-based DNNs. We first analyze the number of unique weights per input neuron of several DNNs. Exploiting common optimizations, such as linear quantization, we observe a very small number of unique weights per input for several FC layers of modern DNNs. Then, to improve the energy-efficiency of FC computation, we present CREW, a hardware accelerator that implements a Computation Reuse and an Efficient Weight Storage mechanism to exploit the large number of repeated weights in FC layers. CREW greatly reduces the number of multiplications and provides significant savings in model memory footprint and memory bandwidth usage. We evaluate CREW on a diverse set of modern DNNs. On average, CREW provides 2.61x speedup and 2.42x energy savings over a TPU-like accelerator.
In third place, we propose a mechanism to optimize the inference of RNNs. RNN cells perform element-wise multiplications across the activations of different gates, sigmoid and tanh being the common activation functions. We perform an analysis of the activation function values, and show that a significant fraction are saturated towards zero or one in popular RNNs. Then, we propose CGPA to dynamically prune activations from RNNs at a coarse granularity. CGPA avoids the evaluation of entire neurons whenever the outputs of peer neurons are saturated. CGPA significantly reduces the amount of computations and memory accesses while avoiding sparsity by a large extent, and can be easily implemented on top of conventional accelerators such as TPU with negligible area overhead, resulting in 12% speedup and 12% energy savings on average for a set of widely used RNNs.
Finally, in the last contribution of this thesis we focus on static DNN pruning methodologies. DNN pruning reduces memory footprint and computational work by removing connections and/or neurons that are ineffectual. However, we show that prior pruning schemes require an extremely time-consuming iterative process that requires retraining the DNN many times to tune the pruning parameters. Then, we propose a DNN pruning scheme based on Principal Component Analysis and relative importance of each neuron's connection that automatically finds the optimized DNN in one shot.Les xarxes neuronals profundes (DNN) han aconseguit un èxit enorme en aplicacions cognitives, i són especialment eficients en problemes de classificació i presa de decisions com ara reconeixement de veu o traducció automàtica. Els dispositius mòbils depenen cada cop més de les DNNs per entendre el món. Els telèfons i rellotges intel·ligents, o fins i tot els cotxes, realitzen diàriament tasques discriminatòries com ara el reconeixement de rostres o objectes. Malgrat la popularitat creixent de les DNNs, el seu funcionament en sistemes mòbils presenta diversos reptes: proporcionar una alta precisió i rendiment amb un petit pressupost de memòria i energia. Les DNNs modernes consisteixen en milions de paràmetres que requereixen recursos computacionals i de memòria enormes i, per tant, no es poden utilitzar directament en sistemes de baixa potència amb recursos limitats. L'objectiu d'aquesta tesi és abordar aquests problemes i proposar noves solucions per tal de dissenyar acceleradors eficients per a sistemes de computació cognitiva basats en DNNs. En primer lloc, ens centrem en optimitzar la inferència de les DNNs per a aplicacions de processament de seqüències. Realitzem una anàlisi de la similitud de les entrades entre execucions consecutives de les DNNs. A continuació, proposem DISC, un accelerador que implementa una tècnica de càlcul diferencial, basat en l'alt grau de semblança de les entrades, per reutilitzar els càlculs de l'execució anterior, en lloc de computar tota la xarxa. Observem que, de mitjana, més del 60% de les entrades de qualsevol capa de les DNNs utilitzades presenten canvis menors respecte a l'execució anterior. Evitar els accessos de memòria i càlculs d'aquestes entrades comporta un estalvi d'energia del 63% de mitjana. En segon lloc, proposem optimitzar la inferència de les DNNs basades en capes FC. Primer analitzem el nombre de pesos únics per neurona d'entrada en diverses xarxes. Aprofitant optimitzacions comunes com la quantització lineal, observem un nombre molt reduït de pesos únics per entrada en diverses capes FC de DNNs modernes. A continuació, per millorar l'eficiència energètica del càlcul de les capes FC, presentem CREW, un accelerador que implementa un eficient mecanisme de reutilització de càlculs i emmagatzematge dels pesos. CREW redueix el nombre de multiplicacions i proporciona estalvis importants en l'ús de la memòria. Avaluem CREW en un conjunt divers de DNNs modernes. CREW proporciona, de mitjana, una millora en rendiment de 2,61x i un estalvi d'energia de 2,42x. En tercer lloc, proposem un mecanisme per optimitzar la inferència de les RNNs. Les cel·les de les xarxes recurrents realitzen multiplicacions element a element de les activacions de diferents comportes, sigmoides i tanh sent les funcions habituals d'activació. Realitzem una anàlisi dels valors de les funcions d'activació i mostrem que una fracció significativa està saturada cap a zero o un en un conjunto d'RNNs populars. A continuació, proposem CGPA per podar dinàmicament les activacions de les RNNs a una granularitat gruixuda. CGPA evita l'avaluació de neurones senceres cada vegada que les sortides de neurones parelles estan saturades. CGPA redueix significativament la quantitat de càlculs i accessos a la memòria, aconseguint en mitjana un 12% de millora en el rendiment i estalvi d'energia. Finalment, en l'última contribució d'aquesta tesi ens centrem en metodologies de poda estàtica de les DNNs. La poda redueix la petjada de memòria i el treball computacional mitjançant l'eliminació de connexions o neurones redundants. Tanmateix, mostrem que els esquemes de poda previs fan servir un procés iteratiu molt llarg que requereix l'entrenament de les DNNs moltes vegades per ajustar els paràmetres de poda. A continuació, proposem un esquema de poda basat en l'anàlisi de components principals i la importància relativa de les connexions de cada neurona que optimitza automàticament el DNN optimitzat en un sol tret sense necessitat de sintonitzar manualment múltiples paràmetres
3
Heath, Felicity. "Variable architecture polymers for DNA delivery." Thesis, University of Nottingham, 2010. http://ethos.bl.uk/OrderDetails.do?uin=uk.bl.ethos.539162.
Повний текст джерела
4
Ding, Ke. "Architectures of DNA block copolymers." [S.l.] : [s.n.], 2006. http://deposit.ddb.de/cgi-bin/dokserv?idn=98214217X.
Повний текст джерела
5
Marriott, Hannah. "Genome architecture and DNA replication in Haloferax volcanii." Thesis, University of Nottingham, 2018. http://eprints.nottingham.ac.uk/50190/.
Повний текст джерелаАнотація:
The archaeon Haloferax volcanii is used to study DNA replication and repair, and it is unique amongst cellular organisms as it is able to grow in the absence of DNA replication origins. There are four DNA replication origins on the main circular chromosome (including the integrated mega-plasmid pHV4) and one on each of the other mega-plasmids pHV1 and pHV3. Replication origins are normally required for the initiation of DNA replication, however H. volcanii is able to grow faster when all chromosomal origins have been deleted. Therefore, H. volcanii must utilise other methods of DNA replication such as recombination-dependent replication. The origin found on pHV3 cannot be deleted from the episomal mega-plasmid, whereas the origin can be deleted from episomal pHV4. The pHV3 mega- plasmid can be integrated onto the main chromosome, which allows the pHV3 origin to be deleted from the chromosome. The pHV1 mega-plasmid origin can be deleted from the episomal mega-plasmid, and the entire mega-plasmid can be lost from the H. volcanii cell. This generates a viable, healthy strain, which shows that the pHV1 mega-plasmid is non- essential. It was also found that the pHV1 mega-plasmid exists in H. volcanii as a 6x concatemer which is ~510 kb in size, which may explain the reason for being able to delete the origin. To further investigate the mechanisms that recombination-dependent replication may use, replication machinery (MCM and GINS) were tagged and expressed. Due to time constraints, interactions were not seen. The mcm gene was put under the control of a tryptophan inducible promoter. A strain lacking chromosomal origins and therefore primarily using recombination-dependent replication was shown to require more MCM than a wild-type strain.
6
Wei, Diming. "The beauty of DNA architecture : the design and applications in DNA nanotechnology /." View abstract or full-text, 2009. http://library.ust.hk/cgi/db/thesis.pl?CBME%202009%20WEI.
Повний текст джерела
7
Schilter, David. "Synthesis and DNA-binding of Metallocyclic Architectures." Thesis, The University of Sydney, 2009. http://hdl.handle.net/2123/5317.
Повний текст джерелаАнотація:
A new family of cationic N-heterocyclic ligand derivatives was prepared and characterised. Among these compounds are halide salts of the dications [Y(spacer)Y]2+, each of which comprise two N heterocyclic donor groups (Y = 4,4′-bipy, pyz, apyz, apym) linked by a conformationally flexible spacer such as (CH2)n, α,α′-xylylene, 2,6-lutidylene or thiabicyclo[3.3.1]nonane-2,6 diyl. The diquaternary halide salts were converted to NO3- and PF6- salts, and interaction of these bridging ligands with labile palladium(II) and platinum(II) precursors afforded several multinuclear complexes. Bis(4,4′-bipyridinium) dications were incorporated into the dinuclear macrocycles [M2(2,2′ bipy)2{4,4′ bipy(CH2)n4,4′-bipy}2]8+ (M = Pd, Pt; n = 4, 6), cis [Pd2Cl4{4,4′ bipy(CH2)34,4′-bipy}2]4+, [Pt2(dppp)2{4,4′-bipy(1,2-xylylene)4,4′-bipy}2]8+ and cis-[Pt2Cl4{4,4′-bipy(1,2-xylylene)4,4′-bipy}2]4+. While bis(pyrazinium) analogues were unreactive towards the palladium(II) and platinum(II) precursors, the doubly deprotonated bis(3 aminopyrazinium) and bis(2 aminopyrimidinium) derivatives served as charge-neutral quadruply-bridging ligands in the complexes [Pt4(2,2′ bipy)4{apyz(CH2)6apyz–2H}2]8+ and [Pt4(2,2′ bipy)4{apym(CH2)5apym–2H}2]8+, both of which feature Pt(II). Pt(II) interactions. Larger species formed when the diamine O,O′-bis(2-aminoethyl)octadeca(ethylene glycol) (PEGda) was treated with cis dinitratopalladium(II) and platinum(II) precursors. The resulting complexes [M(N,N)(PEGda)]2+ (M = Pd, Pt; N,N = 2,2′-bipy, en, tmeda) possessed great size (62 membered chelate rings) and aqueous solubility. DNA-binding studies were conducted with selected complexes in order to investigate the types of interactions these species might participate in. Equimolar mixtures containing either the 16mer duplex DNA D2 or the single strand D2a and palladium(II)/platinum(II) complexes were prepared and analysed by negative-ion ESI MS. Studies of D2/Pd(II) mixtures suggested extensive fragmentation was occuring, and the use of [Pd(tmeda)(PEGda)]2+ and [Pd2(2,2′-bipy)2{4,4′-bipy(CH2)44,4′-bipy}2]8+ resulted in D2 adducts of [Pd(tmeda)]2+ and [4,4′-bipy(CH2)44,4′-bipy]2+, respectively. Decomposition also occurred when D2a was used, although 1 : 1 adducts were observed with [Pd(tmeda)(PEGda)]2+, [Pd(2,2′ bipy)(PEGda)]2+ and [Pd2(2,2′-bipy)2{4,4′-bipy(CH2)44,4′-bipy}2]8+. The low intensities of these adducts indicated that they are unstable towards ESI MS. Analogous ESI-MS experiments using platinum(II) derivatives were performed and, in contrast to those with palladium(II), indicated that the complexes remained largely intact. ESI-MS analysis of D2/Pt(II) mixtures allowed for the detection of 1 : 1 D2 adducts of [Pt(en)(PEGda)]2+, [Pt(tmeda)(PEGda)]2+ and [Pt2(2,2′-bipy)2{4,4′-bipy(CH2)44,4′-bipy}2]8+. Intensities of the adduct ions suggested the greater charge and aryl surface area allow the dinuclear species to bind D2 most strongly. Both [Pt(2,2′-bipy)(Mebipy)2]4+ and [Pt(2,2′ bipy)(NH3)2]2+ gave rise to 1 : 2 adducts of D2, although the latter was found to be a weaker binder, perhaps owing to its lower charge. Data obtained using 1 : 5 (D2 : complex) mixtures were consistent with the results above and suggested that D2 can bind more molecules of daunomycin than any of the platinum(II) species. Analyses of D2a/Pt(II) mixtures gave results similar to those obtained with D2, although fragmentation was more pronounced, indicating that the nucleobases in D2a play more significant roles in mediating decomposition than those in D2, in which they are paired in a complementary manner. Investigations into the effects of selected platinum(II) complexes on the thermal denaturation of calf-thymus DNA (CT-DNA) in solution were conducted. Both [Pt2(2,2′ bipy)2{4,4′-bipy(CH2)64,4′-bipy}2]8+ and [Pt(2,2′-bipy)(Mebipy)2]4+ greatly stabilised CT-DNA, most likely by intercalation. In contrast, [Pt(tmeda)(PEGda)]2+ and [Pt(en)(PEGda)]2+ (as well as PEGda) caused negligible changes in melting temperature (∆Tm), suggesting that these interact weakly with CT-DNA. Data for [Pt(2,2′ bipy)(PEGda)]2+ and [Pt(2,2′-bipy)(NH3)2]2+ indicated that these species perhaps intercalate CT-DNA, with similar ∆Tm values for both complexes implying that PEGda does not play a major role in binding. While findings from ESI-MS experiments were similar to those from the thermal denaturation experiments, discrepancies between results from the two methods could be found. In particular, fragmentation of cyclic species during ESI-MS caused the binding strength of the species to be underestimated when this method was employed.
8
Schilter, David. "Synthesis and DNA-binding of Metallocyclic Architectures." University of Sydney, 2009. http://hdl.handle.net/2123/5317.
Повний текст джерелаАнотація:
PhDA new family of cationic N-heterocyclic ligand derivatives was prepared and characterised. Among these compounds are halide salts of the dications [Y(spacer)Y]2+, each of which comprise two N heterocyclic donor groups (Y = 4,4′-bipy, pyz, apyz, apym) linked by a conformationally flexible spacer such as (CH2)n, α,α′-xylylene, 2,6-lutidylene or thiabicyclo[3.3.1]nonane-2,6 diyl. The diquaternary halide salts were converted to NO3- and PF6- salts, and interaction of these bridging ligands with labile palladium(II) and platinum(II) precursors afforded several multinuclear complexes. Bis(4,4′-bipyridinium) dications were incorporated into the dinuclear macrocycles [M2(2,2′ bipy)2{4,4′ bipy(CH2)n4,4′-bipy}2]8+ (M = Pd, Pt; n = 4, 6), cis [Pd2Cl4{4,4′ bipy(CH2)34,4′-bipy}2]4+, [Pt2(dppp)2{4,4′-bipy(1,2-xylylene)4,4′-bipy}2]8+ and cis-[Pt2Cl4{4,4′-bipy(1,2-xylylene)4,4′-bipy}2]4+. While bis(pyrazinium) analogues were unreactive towards the palladium(II) and platinum(II) precursors, the doubly deprotonated bis(3 aminopyrazinium) and bis(2 aminopyrimidinium) derivatives served as charge-neutral quadruply-bridging ligands in the complexes [Pt4(2,2′ bipy)4{apyz(CH2)6apyz–2H}2]8+ and [Pt4(2,2′ bipy)4{apym(CH2)5apym–2H}2]8+, both of which feature Pt(II). Pt(II) interactions. Larger species formed when the diamine O,O′-bis(2-aminoethyl)octadeca(ethylene glycol) (PEGda) was treated with cis dinitratopalladium(II) and platinum(II) precursors. The resulting complexes [M(N,N)(PEGda)]2+ (M = Pd, Pt; N,N = 2,2′-bipy, en, tmeda) possessed great size (62 membered chelate rings) and aqueous solubility. DNA-binding studies were conducted with selected complexes in order to investigate the types of interactions these species might participate in. Equimolar mixtures containing either the 16mer duplex DNA D2 or the single strand D2a and palladium(II)/platinum(II) complexes were prepared and analysed by negative-ion ESI MS. Studies of D2/Pd(II) mixtures suggested extensive fragmentation was occuring, and the use of [Pd(tmeda)(PEGda)]2+ and [Pd2(2,2′-bipy)2{4,4′-bipy(CH2)44,4′-bipy}2]8+ resulted in D2 adducts of [Pd(tmeda)]2+ and [4,4′-bipy(CH2)44,4′-bipy]2+, respectively. Decomposition also occurred when D2a was used, although 1 : 1 adducts were observed with [Pd(tmeda)(PEGda)]2+, [Pd(2,2′ bipy)(PEGda)]2+ and [Pd2(2,2′-bipy)2{4,4′-bipy(CH2)44,4′-bipy}2]8+. The low intensities of these adducts indicated that they are unstable towards ESI MS. Analogous ESI-MS experiments using platinum(II) derivatives were performed and, in contrast to those with palladium(II), indicated that the complexes remained largely intact. ESI-MS analysis of D2/Pt(II) mixtures allowed for the detection of 1 : 1 D2 adducts of [Pt(en)(PEGda)]2+, [Pt(tmeda)(PEGda)]2+ and [Pt2(2,2′-bipy)2{4,4′-bipy(CH2)44,4′-bipy}2]8+. Intensities of the adduct ions suggested the greater charge and aryl surface area allow the dinuclear species to bind D2 most strongly. Both [Pt(2,2′-bipy)(Mebipy)2]4+ and [Pt(2,2′ bipy)(NH3)2]2+ gave rise to 1 : 2 adducts of D2, although the latter was found to be a weaker binder, perhaps owing to its lower charge. Data obtained using 1 : 5 (D2 : complex) mixtures were consistent with the results above and suggested that D2 can bind more molecules of daunomycin than any of the platinum(II) species. Analyses of D2a/Pt(II) mixtures gave results similar to those obtained with D2, although fragmentation was more pronounced, indicating that the nucleobases in D2a play more significant roles in mediating decomposition than those in D2, in which they are paired in a complementary manner. Investigations into the effects of selected platinum(II) complexes on the thermal denaturation of calf-thymus DNA (CT-DNA) in solution were conducted. Both [Pt2(2,2′ bipy)2{4,4′-bipy(CH2)64,4′-bipy}2]8+ and [Pt(2,2′-bipy)(Mebipy)2]4+ greatly stabilised CT-DNA, most likely by intercalation. In contrast, [Pt(tmeda)(PEGda)]2+ and [Pt(en)(PEGda)]2+ (as well as PEGda) caused negligible changes in melting temperature (∆Tm), suggesting that these interact weakly with CT-DNA. Data for [Pt(2,2′ bipy)(PEGda)]2+ and [Pt(2,2′-bipy)(NH3)2]2+ indicated that these species perhaps intercalate CT-DNA, with similar ∆Tm values for both complexes implying that PEGda does not play a major role in binding. While findings from ESI-MS experiments were similar to those from the thermal denaturation experiments, discrepancies between results from the two methods could be found. In particular, fragmentation of cyclic species during ESI-MS caused the binding strength of the species to be underestimated when this method was employed.
9
Yu, Zhiling. "Interactions and architecture of human MCM proteins in vitro and in vivo /." View Abstract or Full-Text, 2003. http://library.ust.hk/cgi/db/thesis.pl?BICH%202003%20YU.
Повний текст джерелаАнотація:
Thesis (Ph. D.)--Hong Kong University of Science and Technology, 2003.Includes bibliographical references (leaves 118-137). Also available in electronic version. Access restricted to campus users.
10
van, der Merwe Mariè. "Enzyme architecture and flexibility affect DNA topoisomerase I function." View the abstract Download the full-text PDF version, 2007. http://etd.utmem.edu/ABSTRACTS/2007-026-van_der_Merwe-Index.html.
Повний текст джерелаАнотація:
Thesis (Ph.D.)--University of Tennessee Health Science Center, 2007.Title from title page screen (viewed on July 29, 2008). Research advisor: Mary-Ann Bjornsti, Ph.D. Document formatted into pages (xiii, 175 p. : ill.). Vita. Abstract. Includes bibliographical references (p. 161-175).
Книги з теми "DNN architecture"
1
Charre, Alain. Dan Graham. Paris: Dis Voir, 1995.
Знайти повний текст джерела
2
Charre, Alain. Dan Graham. Paris: Editions Dis voir, 1995.
Знайти повний текст джерела
3
1944-, Budihardjo Eko, ed. Pengaruh budaya dan iklim dalam perancangan arsitektur. Bandung: Alumni, 2009.
Знайти повний текст джерела
4
1944-, Budihardjo Eko, ed. Pengaruh budaya dan iklim dalam perancangan arsitektur. Bandung: Alumni, 2009.
Знайти повний текст джерела
5
Tiantian, Xu, ed. Architecture as transformer: DnA-Design and Architecture, Beijing : projects 2004-2018. Berlin: Aedes, 2018.
Знайти повний текст джерела
6
Über den Minderwertigkeitskomplex der deutschen Architektur: Ursachen einer Kontroverse. Göttingen: Optimus Verlag, 2010.
Знайти повний текст джерела
7
Mayr, Fingerle Christoph, ed. Neues Bauen in den Alpen: Architekturpreis 2006 = Architettura alpina contemporanea : premio d'architettura 2006 = New Alpine architecture : architectural prize 2006. Basel: Birkhäuser, 2008.
Знайти повний текст джерела
8
van, Boven Cees, Freijser Victor, and Vaillant Christiaan, eds. Gids van de moderne architectuur in Den Haag =: Guide to modern architecture in The Hague. 2nd ed. Den Haag: Uitgeverij Ulysses, 1998.
Знайти повний текст джерела
9
1931-, Böhm Elisabeth, ed. Gottfried Böhm: Bauten und Projekte : Auszug aus den Jahren 1985-2000 = Buildings and projects : a selection of works 1985-2000. Tübingen: Wasmuth, 2001.
Знайти повний текст джерела
10
Foundation, Solomon R. Guggenheim, ed. Dan Flavin: The architecture of light. New York: The Solomon R. Guggenheim Foundation, 1999.
Знайти повний текст джерелаЧастини книг з теми "DNN architecture"
1
Wang, Liang, and Jianxin Zhao. "Deep Neural Networks." In Architecture of Advanced Numerical Analysis Systems, 121–47. Berkeley, CA: Apress, 2022. http://dx.doi.org/10.1007/978-1-4842-8853-5_5.
Повний текст джерелаАнотація:
AbstractThere are many articles teaching people how to build intelligent applications using different frameworks such as TensorFlow, PyTorch, etc. However, except those very professional research papers, very few articles can give us a comprehensive understanding on how to develop such frameworks. In this chapter, rather than just “casting spells,” we focus on explaining how to make the magic work in the first place. We will dissect the deep neural network module in Owl, then demonstrate how to assemble different building blocks to build a working framework. Owl’s neural network module is a full-featured DNN framework. You can define a neural network in a very compact and elegant way thanks to OCaml’s expressiveness. The DNN applications built on Owl can achieve state-of-the-art performance.
2
Bartz-Beielstein, Thomas, Sowmya Chandrasekaran, and Frederik Rehbach. "Case Study III: Tuning of Deep Neural Networks." In Hyperparameter Tuning for Machine and Deep Learning with R, 235–69. Singapore: Springer Nature Singapore, 2023. http://dx.doi.org/10.1007/978-981-19-5170-1_10.
Повний текст джерелаАнотація:
AbstractA surrogate model based Hyperparameter Tuning (HPT) approach for Deep Learning (DL) is presented. This chapter demonstrates how the architecture-level parameters (hyperparameters) of Deep Neural Networks (DNNs) that were implemented in / can be optimized. The implementation of the tuning procedure is 100% accessible from R, the software environment for statistical computing. How the software packages (, , and ) can be combined in a very efficient and effective manner will be exemplified in this chapter. The hyperparameters of a standard DNN are tuned. The performances of the six Machine Learning (ML) methods discussed in this book are compared to the results from the DNN. This study provides valuable insights in the tunability of several methods, which is of great importance for the practitioner.
3
Han, Donghyeon, and Hoi-Jun Yoo. "An Overview of Energy-Efficient DNN Training Processors." In On-Chip Training NPU - Algorithm, Architecture and SoC Design, 183–210. Cham: Springer Nature Switzerland, 2023. http://dx.doi.org/10.1007/978-3-031-34237-0_8.
Повний текст джерела
4
Anjum, Muhammad, Moizzah Asif, and Jonathan Williams. "Towards an Optimal Deep Neural Network for SOC Estimation of Electric-Vehicle Lithium-Ion Battery Cells." In Springer Proceedings in Energy, 11–18. Cham: Springer International Publishing, 2021. http://dx.doi.org/10.1007/978-3-030-63916-7_2.
Повний текст джерелаАнотація:
AbstractThis paper has identified a minimal configuration of a DNN architecture and hyperparameter settings to effectively estimate SOC of EV battery cells. The results from the experimental work has shown that a minimal configuration of hidden layers and neurons can reduce the computational cost and resources required without compromising the performance. This is further supported by the number of epochs taken to train the best DNN SOC estimations model. Hence, demonstrating that, the risk of overfitting estimation models to training datasets, can also be subsided. This is further supported by the generalisation capability of the best model demonstrated through the decrease in error metrics values from test phase to those in validation phase.
5
Pontes, Felipe Arruda, and Edward Curry. "Cloud-Edge Microservice Architecture for DNN-based Distributed Multimedia Event Processing." In Communications in Computer and Information Science, 65–72. Cham: Springer International Publishing, 2021. http://dx.doi.org/10.1007/978-3-030-71906-7_6.
Повний текст джерела
6
Han, Donghyeon, and Hoi-Jun Yoo. "HNPU-V1: An Adaptive DNN Training Processor Utilizing Stochastic Dynamic Fixed-Point and Active Bit-Precision Searching." In On-Chip Training NPU - Algorithm, Architecture and SoC Design, 121–61. Cham: Springer Nature Switzerland, 2023. http://dx.doi.org/10.1007/978-3-031-34237-0_6.
Повний текст джерела
7
Han, Donghyeon, and Hoi-Jun Yoo. "HNPU-V2: An Energy-Efficient DNN Training Processor for Robust Object Detection with Real-World Environmental Adaptation." In On-Chip Training NPU - Algorithm, Architecture and SoC Design, 163–82. Cham: Springer Nature Switzerland, 2023. http://dx.doi.org/10.1007/978-3-031-34237-0_7.
Повний текст джерела
8
Sinha, Bam Bahadur, Gurvinder Singh Yadav, and Sagar Badrish Kudkelwar. "Modified-PIP with Deep Neural Network (DNN) Architecture: A Coherent Recommendation Framework for Capturing User Behaviour." In Studies in Big Data, 121–40. Cham: Springer International Publishing, 2022. http://dx.doi.org/10.1007/978-3-031-10869-3_7.
Повний текст джерела
9
Alsuhli, Ghada, Vasilis Sakellariou, Hani Saleh, Mahmoud Al-Qutayri, Baker Mohammad, and Thanos Stouraitis. "BFP for DNN Architectures." In Synthesis Lectures on Engineering, Science, and Technology, 61–72. Cham: Springer Nature Switzerland, 2023. http://dx.doi.org/10.1007/978-3-031-38133-1_6.
Повний текст джерела
10
Alsuhli, Ghada, Vasilis Sakellariou, Hani Saleh, Mahmoud Al-Qutayri, Baker Mohammad, and Thanos Stouraitis. "Posit for DNN Architectures." In Synthesis Lectures on Engineering, Science, and Technology, 81–88. Cham: Springer Nature Switzerland, 2023. http://dx.doi.org/10.1007/978-3-031-38133-1_8.
Повний текст джерелаТези доповідей конференцій з теми "DNN architecture"
1
Krishnan, Gokul, Zhenyu Wang, Li Yang, Injune Yeo, Jian Meng, Rajiv V. Joshi, Nathaniel C. Cady, Deliang Fan, Jae-Sun Seo, and Yu Cao. "IMC Architecture for Robust DNN Acceleration." In 2022 IEEE 16th International Conference on Solid-State & Integrated Circuit Technology (ICSICT). IEEE, 2022. http://dx.doi.org/10.1109/icsict55466.2022.9963165.
Повний текст джерела
2
Yemini, Yochai, Shlomo E. Chazan, Jacob Goldberger, and Sharon Gannot. "A Composite DNN Architecture for Speech Enhancement." In ICASSP 2020 - 2020 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP). IEEE, 2020. http://dx.doi.org/10.1109/icassp40776.2020.9053821.
Повний текст джерела
3
Kim, Jinhan, Nargiz Humbatova, Gunel Jahangirova, Paolo Tonella, and Shin Yoo. "Repairing DNN Architecture: Are We There Yet?" In 2023 IEEE Conference on Software Testing, Verification and Validation (ICST). IEEE, 2023. http://dx.doi.org/10.1109/icst57152.2023.00030.
Повний текст джерела
4
He, Zhezhi. "Session details: Architecture for DNN Acceleration (Virtual)." In ICCAD '22: IEEE/ACM International Conference on Computer-Aided Design. New York, NY, USA: ACM, 2022. http://dx.doi.org/10.1145/3578439.
Повний текст джерела
5
Yazdani, Reza, Marc Riera, Jose-Maria Arnau, and Antonio Gonzalez. "The Dark Side of DNN Pruning." In 2018 ACM/IEEE 45th Annual International Symposium on Computer Architecture (ISCA). IEEE, 2018. http://dx.doi.org/10.1109/isca.2018.00071.
Повний текст джерела
6
Kwon, Hyoukjun, Liangzhen Lai, Michael Pellauer, Tushar Krishna, Yu-Hsin Chen, and Vikas Chandra. "Heterogeneous Dataflow Accelerators for Multi-DNN Workloads." In 2021 IEEE International Symposium on High-Performance Computer Architecture (HPCA). IEEE, 2021. http://dx.doi.org/10.1109/hpca51647.2021.00016.
Повний текст джерела
7
Cai, Chengtao, and Dongning Guo. "CNN-Self-Attention-DNN Architecture For Mandarin Recognition." In 2020 Chinese Control And Decision Conference (CCDC). IEEE, 2020. http://dx.doi.org/10.1109/ccdc49329.2020.9164333.
Повний текст джерела
8
Geng, Wei, Dongyu Liu, and Xiu Cao. "A Power Anomaly Detection Architecture Based on DNN." In the 3rd International Conference. New York, New York, USA: ACM Press, 2019. http://dx.doi.org/10.1145/3331453.3361641.
Повний текст джерела
9
Glint, Tom, Chandan Kumar Jha, Manu Awasthi, and Joycee Mekie. "Analysis of Quantization Across DNN Accelerator Architecture Paradigms." In 2023 Design, Automation & Test in Europe Conference & Exhibition (DATE). IEEE, 2023. http://dx.doi.org/10.23919/date56975.2023.10136899.
Повний текст джерела
10
Janfaza, Vahid, Kevin Weston, Moein Razavi, Shantanu Mandal, Farabi Mahmud, Alex Hilty, and Abdullah Muzahid. "MERCURY: Accelerating DNN Training By Exploiting Input Similarity." In 2023 IEEE International Symposium on High-Performance Computer Architecture (HPCA). IEEE, 2023. http://dx.doi.org/10.1109/hpca56546.2023.10071051.
Повний текст джерелаЗвіти організацій з теми "DNN architecture"
1
Tayeb, Shahab. Taming the Data in the Internet of Vehicles. Mineta Transportation Institute, January 2022. http://dx.doi.org/10.31979/mti.2022.2014.
Повний текст джерелаАнотація:
As an emerging field, the Internet of Vehicles (IoV) has a myriad of security vulnerabilities that must be addressed to protect system integrity. To stay ahead of novel attacks, cybersecurity professionals are developing new software and systems using machine learning techniques. Neural network architectures improve such systems, including Intrusion Detection System (IDSs), by implementing anomaly detection, which differentiates benign data packets from malicious ones. For an IDS to best predict anomalies, the model is trained on data that is typically pre-processed through normalization and feature selection/reduction. These pre-processing techniques play an important role in training a neural network to optimize its performance. This research studies the impact of applying normalization techniques as a pre-processing step to learning, as used by the IDSs. The impacts of pre-processing techniques play an important role in training neural networks to optimize its performance. This report proposes a Deep Neural Network (DNN) model with two hidden layers for IDS architecture and compares two commonly used normalization pre-processing techniques. Our findings are evaluated using accuracy, Area Under Curve (AUC), Receiver Operator Characteristic (ROC), F-1 Score, and loss. The experimentations demonstrate that Z-Score outperforms no-normalization and the use of Min-Max normalization.
2
Benner, Steven A. Design Automation Software for DNA-Based Nano-Sensor Architecture. Fort Belvoir, VA: Defense Technical Information Center, April 2012. http://dx.doi.org/10.21236/ada582334.
Повний текст джерела
3
Clark, Paul C., Timothy E. Lavin, Cynthia E. Irvine, and David J. Shifflett. DNS and Multilevel Secure Networks: Architectures and Recommendations. Fort Belvoir, VA: Defense Technical Information Center, February 2009. http://dx.doi.org/10.21236/ada498511.
Повний текст джерела
4
Peterson, J., O. Kolkman, H. Tschofenig, and B. Aboba. Architectural Considerations on Application Features in the DNS. RFC Editor, October 2013. http://dx.doi.org/10.17487/rfc6950.
Повний текст джерела
5
Macula, Anthony, Russell Deaton, and Junghuei Chen. A Two-Dimensional Deoxyribonucleic Acid (DNA) Matrix Based Biomolecular Computing and Memory Architecture. Fort Belvoir, VA: Defense Technical Information Center, February 2009. http://dx.doi.org/10.21236/ada494650.
Повний текст джерела
6
Yu, Haichao, Haoxiang Li, Honghui Shi, Thomas S. Huang, and Gang Hua. Any-Precision Deep Neural Networks. Web of Open Science, December 2020. http://dx.doi.org/10.37686/ejai.v1i1.82.
Повний текст джерелаАнотація:
We present Any-Precision Deep Neural Networks (Any- Precision DNNs), which are trained with a new method that empowers learned DNNs to be flexible in any numerical precision during inference. The same model in runtime can be flexibly and directly set to different bit-width, by trun- cating the least significant bits, to support dynamic speed and accuracy trade-off. When all layers are set to low- bits, we show that the model achieved accuracy compara- ble to dedicated models trained at the same precision. This nice property facilitates flexible deployment of deep learn- ing models in real-world applications, where in practice trade-offs between model accuracy and runtime efficiency are often sought. Previous literature presents solutions to train models at each individual fixed efficiency/accuracy trade-off point. But how to produce a model flexible in runtime precision is largely unexplored. When the demand of efficiency/accuracy trade-off varies from time to time or even dynamically changes in runtime, it is infeasible to re-train models accordingly, and the storage budget may forbid keeping multiple models. Our proposed framework achieves this flexibility without performance degradation. More importantly, we demonstrate that this achievement is agnostic to model architectures. We experimentally validated our method with different deep network backbones (AlexNet-small, Resnet-20, Resnet-50) on different datasets (SVHN, Cifar-10, ImageNet) and observed consistent results.
7
Joel, Daniel M., Steven J. Knapp, and Yaakov Tadmor. Genomic Approaches for Understanding Virulence and Resistance in the Sunflower-Orobanche Host-Parasite Interaction. United States Department of Agriculture, August 2011. http://dx.doi.org/10.32747/2011.7592655.bard.
Повний текст джерелаАнотація:
Oroginal Objectives: (i) identify DNA markers linked to the avirulence (Avr) locus and locate the Avr locus through genetic mapping with an inter-race Orobanche cumana population; (ii) develop high-throughput fingerprint DNA markers for genotypingO. cumana races; (iii) identify nucleotide binding domain leucine rich repeat (NB-LRR) genes encoding R proteins conferring resistance to O. cumana in sunflower; (iv) increase the resolution of the chromosomal segment harboring Or₅ and related R genes through genetic and physical mapping in previously and newly developed mapping populations of sunflower; and (v) develop high-throughput DNA markers for rapidly and efficiently identifying and transferring sunflower R genes through marker-assisted selection. Revisions made during the course of project: Following changes in O. cumana race distribution in Israel, the newly arrived virulent race H was chosen for further analysis. HA412-HO, which was primarily chosen as a susceptible sunflower cultivar, was more resistant to the new parasite populations than var. Shemesh, thus we shifted sunflower research into analyzing the resistance of HA412-HO. We exceeded the deliverables for Objectives #3-5 by securing funding for complete physical and high-density genetic mapping of the sunflower genome, in addition to producing a complete draft sequence of the sunflower genome. We discovered limited diversity between the parents of the O. cumana population developed for the mapping study. Hence, the developed DNA marker resources were insufficient to support genetic map construction. This objective was beyond the scale and scope of the funding. This objective is challenging enough to be the entire focus of follow up studies. Background to the topic: O. cumana, an obligate parasitic weed, is one of the most economically important and damaging diseases of sunflower, causes significant yield losses in susceptible genotypes, and threatens production in Israel and many other countries. Breeding for resistance has been crucial for protecting sunflower from O. cumana, and problematic because new races of the pathogen continually emerge, necessitating discovery and deployment of new R genes. The process is challenging because of the uncertainty in identifying races in a genetically diverse parasite. Major conclusions, solutions, achievements: We developed a small collection of SSR markers for genetic mapping in O. cumana and completed a diversity study to lay the ground for objective #1. Because DNA sequencing and SNPgenotyping technology dramatically advanced during the course of the study, we recommend shifting future work to SNP discovery and mapping using array-based approaches, instead of SSR markers. We completed a pilot study using a 96-SNP array, but it was not large enough to support genetic mapping in O.cumana. The development of further SNPs was beyond the scope of the grant. However, the collection of SSR markers was ideal for genetic diversity analysis, which indicated that O. cumanapopulations in Israel considerably differ frompopulations in other Mediterranean countries. We supplied physical and genetic mapping resources for identifying R-genes in sunflower responsible for resistance to O. cumana. Several thousand mapped SNP markers and a complete draft of the sunflower genome sequence are powerful tools for identifying additional candidate genes and understanding the genomic architecture of O. cumana-resistanceanddisease-resistance genes. Implications: The OrobancheSSR markers have utility in sunflower breeding and genetics programs, as well as a tool for understanding the heterogeneity of races in the field and for geographically mapping of pathotypes.The segregating populations of both Orobanche and sunflower hybrids are now available for QTL analyses.
8
Crowley, David E., Dror Minz, and Yitzhak Hadar. Shaping Plant Beneficial Rhizosphere Communities. United States Department of Agriculture, July 2013. http://dx.doi.org/10.32747/2013.7594387.bard.
Повний текст джерелаАнотація:
PGPR bacteria include taxonomically diverse bacterial species that function for improving plant mineral nutrition, stress tolerance, and disease suppression. A number of PGPR are being developed and commercialized as soil and seed inoculants, but to date, their interactions with resident bacterial populations are still poorly understood, and-almost nothing is known about the effects of soil management practices on their population size and activities. To this end, the original objectives of this research project were: 1) To examine microbial community interactions with plant-growth-promoting rhizobacteria (PGPR) and their plant hosts. 2) To explore the factors that affect PGPR population size and activity on plant root surfaces. In our original proposal, we initially prqposed the use oflow-resolution methods mainly involving the use of PCR-DGGE and PLFA profiles of community structure. However, early in the project we recognized that the methods for studying soil microbial communities were undergoing an exponential leap forward to much more high resolution methods using high-throughput sequencing. The application of these methods for studies on rhizosphere ecology thus became a central theme in these research project. Other related research by the US team focused on identifying PGPR bacterial strains and examining their effective population si~es that are required to enhance plant growth and on developing a simulation model that examines the process of root colonization. As summarized in the following report, we characterized the rhizosphere microbiome of four host plant species to determine the impact of the host (host signature effect) on resident versus active communities. Results of our studies showed a distinct plant host specific signature among wheat, maize, tomato and cucumber, based on the following three parameters: (I) each plant promoted the activity of a unique suite of soil bacterial populations; (2) significant variations were observed in the number and the degree of dominance of active populations; and (3)the level of contribution of active (rRNA-based) populations to the resident (DNA-based) community profiles. In the rhizoplane of all four plants a significant reduction of diversity was observed, relative to the bulk soil. Moreover, an increase in DNA-RNA correspondence indicated higher representation of active bacterial populations in the residing rhizoplane community. This research demonstrates that the host plant determines the bacterial community composition in its immediate vicinity, especially with respect to the active populations. Based on the studies from the US team, we suggest that the effective population size PGPR should be maintained at approximately 105 cells per gram of rhizosphere soil in the zone of elongation to obtain plant growth promotion effects, but emphasize that it is critical to also consider differences in the activity based on DNA-RNA correspondence. The results ofthis research provide fundamental new insight into the composition ofthe bacterial communities associated with plant roots, and the factors that affect their abundance and activity on root surfaces. Virtually all PGPR are multifunctional and may be expected to have diverse levels of activity with respect to production of plant growth hormones (regulation of root growth and architecture), suppression of stress ethylene (increased tolerance to drought and salinity), production of siderophores and antibiotics (disease suppression), and solubilization of phosphorus. The application of transcriptome methods pioneered in our research will ultimately lead to better understanding of how management practices such as use of compost and soil inoculants can be used to improve plant yields, stress tolerance, and disease resistance. As we look to the future, the use of metagenomic techniques combined with quantitative methods including microarrays, and quantitative peR methods that target specific genes should allow us to better classify, monitor, and manage the plant rhizosphere to improve crop yields in agricultural ecosystems. In addition, expression of several genes in rhizospheres of both cucumber and whet roots were identified, including mostly housekeeping genes. Denitrification, chemotaxis and motility genes were preferentially expressed in wheat while in cucumber roots bacterial genes involved in catalase, a large set of polysaccharide degradation and assimilatory sulfate reduction genes were preferentially expressed.
9
Weller, Joel I., Derek M. Bickhart, Micha Ron, Eyal Seroussi, George Liu, and George R. Wiggans. Determination of actual polymorphisms responsible for economic trait variation in dairy cattle. United States Department of Agriculture, January 2015. http://dx.doi.org/10.32747/2015.7600017.bard.
Повний текст джерелаАнотація:
The project’s general objectives were to determine specific polymorphisms at the DNA level responsible for observed quantitative trait loci (QTLs) and to estimate their effects, frequencies, and selection potential in the Holstein dairy cattle breed. The specific objectives were to (1) localize the causative polymorphisms to small chromosomal segments based on analysis of 52 U.S. Holstein bulls each with at least 100 sons with high-reliability genetic evaluations using the a posteriori granddaughter design; (2) sequence the complete genomes of at least 40 of those bulls to 20 coverage; (3) determine causative polymorphisms based on concordance between the bulls’ genotypes for specific polymorphisms and their status for a QTL; (4) validate putative quantitative trait variants by genotyping a sample of Israeli Holstein cows; and (5) perform gene expression analysis using statistical methodologies, including determination of signatures of selection, based on somatic cells of cows that are homozygous for contrasting quantitative trait variants; and (6) analyze genes with putative quantitative trait variants using data mining techniques. Current methods for genomic evaluation are based on population-wide linkage disequilibrium between markers and actual alleles that affect traits of interest. Those methods have approximately doubled the rate of genetic gain for most traits in the U.S. Holstein population. With determination of causative polymorphisms, increasing the accuracy of genomic evaluations should be possible by including those genotypes as fixed effects in the analysis models. Determination of causative polymorphisms should also yield useful information on gene function and genetic architecture of complex traits. Concordance between QTL genotype as determined by the a posteriori granddaughter design and marker genotype was determined for 30 trait-by-chromosomal segment effects that are segregating in the U.S. Holstein population; a probability of <10²⁰ was used to accept the null hypothesis that no segregating gene within the chromosomal segment was affecting the trait. Genotypes for 83 grandsires and 17,217 sons were determined by either complete sequence or imputation for 3,148,506 polymorphisms across the entire genome. Variant sites were identified from previous studies (such as the 1000 Bull Genomes Project) and from DNA sequencing of bulls unique to this project, which is one of the largest marker variant surveys conducted for the Holstein breed of cattle. Effects for stature on chromosome 11, daughter pregnancy rate on chromosome 18, and protein percentage on chromosome 20 met 3 criteria: (1) complete or nearly complete concordance, (2) nominal significance of the polymorphism effect after correction for all other polymorphisms, and (3) marker coefficient of determination >40% of total multiple-regression coefficient of determination for the 30 polymorphisms with highest concordance. The missense polymorphism Phe279Tyr in GHR at 31,909,478 base pairs on chromosome 20 was confirmed as the causative mutation for fat and protein concentration. For effect on fat percentage, 12 additional missensepolymorphisms on chromosome 14 were found that had nearly complete concordance with the suggested causative polymorphism (missense mutation Ala232Glu in DGAT1). The markers used in routine U.S. genomic evaluations were increased from 60,000 to 80,000 by adding markers for known QTLs and markers detected in BARD and other research projects. Objectives 1 and 2 were completely accomplished, and objective 3 was partially accomplished. Because no new clear-cut causative polymorphisms were discovered, objectives 4 through 6 were not completed.
10
Eshed-Williams, Leor, and Daniel Zilberman. Genetic and cellular networks regulating cell fate at the shoot apical meristem. United States Department of Agriculture, January 2014. http://dx.doi.org/10.32747/2014.7699862.bard.
Повний текст джерелаАнотація:
The shoot apical meristem establishes plant architecture by continuously producing new lateral organs such as leaves, axillary meristems and flowers throughout the plant life cycle. This unique capacity is achieved by a group of self-renewing pluripotent stem cells that give rise to founder cells, which can differentiate into multiple cell and tissue types in response to environmental and developmental cues. Cell fate specification at the shoot apical meristem is programmed primarily by transcription factors acting in a complex gene regulatory network. In this project we proposed to provide significant understanding of meristem maintenance and cell fate specification by studying four transcription factors acting at the meristem. Our original aim was to identify the direct target genes of WUS, STM, KNAT6 and CNA transcription factor in a genome wide scale and the manner by which they regulate their targets. Our goal was to integrate this data into a regulatory model of cell fate specification in the SAM and to identify key genes within the model for further study. We have generated transgenic plants carrying the four TF with two different tags and preformed chromatin Immunoprecipitation (ChIP) assay to identify the TF direct target genes. Due to unforeseen obstacles we have been delayed in achieving this aim but hope to accomplish it soon. Using the GR inducible system, genetic approach and transcriptome analysis [mRNA-seq] we provided a new look at meristem activity and its regulation of morphogenesis and phyllotaxy and propose a coherent framework for the role of many factors acting in meristem development and maintenance. We provided evidence for 3 different mechanisms for the regulation of WUS expression, DNA methylation, a second receptor pathway - the ERECTA receptor and the CNA TF that negatively regulates WUS expression in its own domain, the Organizing Center. We found that once the WUS expression level surpasses a certain threshold it alters cell identity at the periphery of the inflorescence meristem from floral meristem to carpel fate [FM]. When WUS expression highly elevated in the FM, the meristem turn into indeterminate. We showed that WUS activate cytokinine, inhibit auxin response and represses the genes required for root identity fate and that gradual increase in WUCHEL activity leads to gradual meristem enlargement that affect phyllotaxis. We also propose a model in which the direction of WUS domain expansion laterally or upward affects meristem structure differently. We preformed mRNA-seq on meristems with different size and structure followed by k-means clustering and identified groups of genes that are expressed in specific domains at the meristem. We will integrate this data with the ChIP-seq of the 4 TF to add another layer to the genetic network regulating meristem activity.