Dissertations / Theses on the topic 'Support Vector Machine'
To see the other types of publications on this topic, follow the link: Support Vector Machine.
Create a spot-on reference in APA, MLA, Chicago, Harvard, and other styles
Consult the top 50 dissertations / theses for your research on the topic 'Support Vector Machine.'
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 dissertations / theses on a wide variety of disciplines and organise your bibliography correctly.
1
Cardamone, Dario. "Support Vector Machine a Machine Learning Algorithm." Master's thesis, Alma Mater Studiorum - Università di Bologna, 2017.
Find full textAbstract:
Nella presente tesi di laurea viene preso in considerazione l’algoritmo di classificazione Support Vector Machine. Piu` in particolare si considera la sua formulazione come problema di ottimizazione Mixed Integer Program per la classificazione binaria super- visionata di un set di dati.
2
McChesney, Charlie. "External Support Vector Machine Clustering." ScholarWorks@UNO, 2006. http://scholarworks.uno.edu/td/409.
Full textAbstract:
The external-Support Vector Machine (SVM) clustering algorithm clusters data vectors with no a priori knowledge of each vector's class. The algorithm works by first running a binary SVM against a data set, with each vector in the set randomly labeled, until the SVM converges. It then relabels data points that are mislabeled and a large distance from the SVM hyperplane. The SVM is then iteratively rerun followed by more label swapping until no more progress can be made. After this process, a high percentage of the previously unknown class labels of the data set will be known. With sub-cluster identification upon iterating the overall algorithm on the positive and negative clusters identified (until the clusters are no longer separable into sub-clusters), this method provides a way to cluster data sets without prior knowledge of the data's clustering characteristics, or the number of clusters.
3
Armond, Kenneth C. Jr. "Distributed Support Vector Machine Learning." ScholarWorks@UNO, 2008. http://scholarworks.uno.edu/td/711.
Full textAbstract:
Support Vector Machines (SVMs) are used for a growing number of applications. A fundamental constraint on SVM learning is the management of the training set. This is because the order of computations goes as the square of the size of the training set. Typically, training sets of 1000 (500 positives and 500 negatives, for example) can be managed on a PC without hard-drive thrashing. Training sets of 10,000 however, simply cannot be managed with PC-based resources. For this reason most SVM implementations must contend with some kind of chunking process to train parts of the data at a time (10 chunks of 1000, for example, to learn the 10,000). Sequential and multi-threaded chunking methods provide a way to run the SVM on large datasets while retaining accuracy. The multi-threaded distributed SVM described in this thesis is implemented using Java RMI, and has been developed to run on a network of multi-core/multi-processor computers.
4
Zigic, Ljiljana. "Direct L2 Support Vector Machine." VCU Scholars Compass, 2016. http://scholarscompass.vcu.edu/etd/4274.
Full textAbstract:
This dissertation introduces a novel model for solving the L2 support vector machine dubbed Direct L2 Support Vector Machine (DL2 SVM). DL2 SVM represents a new classification model that transforms the SVM's underlying quadratic programming problem into a system of linear equations with nonnegativity constraints. The devised system of linear equations has a symmetric positive definite matrix and a solution vector has to be nonnegative.
Furthermore, this dissertation introduces a novel algorithm dubbed Non-Negative Iterative Single Data Algorithm (NN ISDA) which solves the underlying DL2 SVM's constrained system of equations. This solver shows significant speedup compared to several other state-of-the-art algorithms. The training time improvement is achieved at no cost, in other words, the accuracy is kept at the same level. All the experiments that support this claim were conducted on various datasets within the strict double cross-validation scheme. DL2 SVM solved with NN ISDA has faster training time on both medium and large datasets.
In addition to a comprehensive DL2 SVM model we introduce and derive its three variants. Three different solvers for the DL2's system of linear equations with nonnegativity constraints were implemented, presented and compared in this dissertation.
5
Park, Yongwon Baskiyar Sanjeev. "Dynamic task scheduling onto heterogeneous machines using Support Vector Machine." Auburn, Ala, 2008. http://repo.lib.auburn.edu/EtdRoot/2008/SPRING/Computer_Science_and_Software_Engineering/Thesis/Park_Yong_50.pdf.
Full text
6
Tsang, Wai-Hung. "Scaling up support vector machines /." View abstract or full-text, 2007. http://library.ust.hk/cgi/db/thesis.pl?CSED%202007%20TSANG.
Full text
7
Perez, Daniel Antonio. "Performance comparison of support vector machine and relevance vector machine classifiers for functional MRI data." Thesis, Georgia Institute of Technology, 2010. http://hdl.handle.net/1853/34858.
Full textAbstract:
Multivariate pattern analysis (MVPA) of fMRI data has been growing in popularity due to its sensitivity to networks of brain activation. It is performed in a predictive modeling framework which is natural for implementing brain state prediction and real-time fMRI applications such as brain computer interfaces. Support vector machines (SVM) have been particularly popular for MVPA owing to their high prediction accuracy even with noisy datasets. Recent work has proposed the use of relevance vector machines (RVM) as an alternative to SVM. RVMs are particularly attractive in time sensitive applications such as real-time fMRI since they tend to perform classification faster than SVMs. Despite the use of both methods in fMRI research, little has been done to compare the performance of these two techniques. This study compares RVM to SVM in terms of time and accuracy to determine which is better suited to real-time applications.
8
Wen, Tong 1970. "Support Vector Machine algorithms : analysis and applications." Thesis, Massachusetts Institute of Technology, 2002. http://hdl.handle.net/1721.1/8404.
Full textAbstract:
Thesis (Ph.D.)--Massachusetts Institute of Technology, Dept. of Mathematics, 2002.Includes bibliographical references (p. 89-97).
Support Vector Machines (SVMs) have attracted recent attention as a learning technique to attack classification problems. The goal of my thesis work is to improve computational algorithms as well as the mathematical understanding of SVMs, so that they can be easily applied to real problems. SVMs solve classification problems by learning from training examples. From the geometry, it is easy to formulate the finding of SVM classifiers as a linearly constrained Quadratic Programming (QP) problem. However, in practice its dual problem is actually computed. An important property of the dual QP problem is that its solution is sparse. The training examples that determine the SVM classifier are known as support vectors (SVs). Motivated by the geometric derivation of the primal QP problem, we investigate how the dual problem is related to the geometry of SVs. This investigation leads to a geometric interpretation of the scaling property of SVMs and an algorithm to further compress the SVs. A random model for the training examples connects the Hessian matrix of the dual QP problem to Wishart matrices. After deriving the distributions of the elements of the inverse Wishart matrix Wn-1(n, nI), we give a conjecture about the summation of the elements of Wn-1(n, nI). It becomes challenging to solve the dual QP problem when the training set is large. We develop a fast algorithm for solving this problem. Numerical experiments show that the MATLAB implementation of this projected Conjugate Gradient algorithm is competitive with benchmark C/C++ codes such as SVMlight and SvmFu. Furthermore, we apply SVMs to time series data.
(cont.) In this application, SVMs are used to predict the movement of the stock market. Our results show that using SVMs has the potential to outperform the solution based on the most widely used geometric Brownian motion model of stock prices.
by Tong Wen.
Ph.D.
9
Liu, Yufeng. "Multicategory psi-learning and support vector machine." Connect to this title online, 2004. http://rave.ohiolink.edu/etdc/view?acc%5Fnum=osu1085424065.
Full textAbstract:
Thesis (Ph. D.)--Ohio State University, 2004.Title from first page of PDF file. Document formatted into pages; contains x, 71 p.; also includes graphics Includes bibliographical references (p. 69-71). Available online via OhioLINK's ETD Center
10
Merat, Sepehr. "Clustering Via Supervised Support Vector Machines." ScholarWorks@UNO, 2008. http://scholarworks.uno.edu/td/857.
Full textAbstract:
An SVM-based clustering algorithm is introduced that clusters data with no a priori knowledge of input classes. The algorithm initializes by first running a binary SVM classifier against a data set with each vector in the set randomly labeled. Once this initialization step is complete, the SVM confidence parameters for classification on each of the training instances can be accessed. The lowest confidence data (e.g., the worst of the mislabeled data) then has its labels switched to the other class label. The SVM is then re-run on the data set (with partly re-labeled data). The repetition of the above process improves the separability until there is no misclassification. Variations on this type of clustering approach are shown.
11
Karlbom, Hannes. "Hybrid Machine Translation : Choosing the best translation with Support Vector Machines." Thesis, Uppsala universitet, Institutionen för informationsteknologi, 2016. http://urn.kb.se/resolve?urn=urn:nbn:se:uu:diva-304257.
Full textAbstract:
In the field of machine translation there are various systems available which have different strengths and weaknesses. This thesis investigates the combination of two systems, a rule based one and a statistical one, to see if such a hybrid system can provide higher quality translations. The classification approach was taken, where a support vector machine is used to choose which sentences from each of the two systems result in the best translation. To label the sentences from the collected data a new method of simulated annealing was applied and compared to previously tried heuristics. The results show that a hybrid system has an increased average BLEU score of 6.10% or 1.86 points over the single best system, and that using the labels created through simulated annealing, over heuristic rules, gives a significant improvement in classifier performance.
12
Wu, Zhili. "Regularization methods for support vector machines." HKBU Institutional Repository, 2008. http://repository.hkbu.edu.hk/etd_ra/912.
Full text
13
Weston, Jason Aaron Edward. "Extensions to the support vector method." Thesis, Royal Holloway, University of London, 2000. http://ethos.bl.uk/OrderDetails.do?uin=uk.bl.ethos.367838.
Full text
14
Zhang, Hang. "Distributed Support Vector Machine With Graphics Processing Units." ScholarWorks@UNO, 2009. http://scholarworks.uno.edu/td/991.
Full textAbstract:
Training a Support Vector Machine (SVM) requires the solution of a very large quadratic programming (QP) optimization problem. Sequential Minimal Optimization (SMO) is a decomposition-based algorithm which breaks this large QP problem into a series of smallest possible QP problems. However, it still costs O(n2) computation time. In our SVM implementation, we can do training with huge data sets in a distributed manner (by breaking the dataset into chunks, then using Message Passing Interface (MPI) to distribute each chunk to a different machine and processing SVM training within each chunk). In addition, we moved the kernel calculation part in SVM classification to a graphics processing unit (GPU) which has zero scheduling overhead to create concurrent threads. In this thesis, we will take advantage of this GPU architecture to improve the classification performance of SVM.
15
Sandoval, Rodríguez Rodrigo Antonio. "Metodología de clasificación dinámica utilizando Support Vector Machine." Tesis, Universidad de Chile, 2007. http://www.repositorio.uchile.cl/handle/2250/102921.
Full textAbstract:
Esta investigación se centra en el problema de clasificación, por medio de confeccionar una metodología que permita detectar y modelar cambios en los patrones que definen la clasificación en el tiempo, en otras palabras, clasificación dinámica.
La metodología desarrollada propone utilizar los resultados obtenidos en un periodo de tiempo para la construcción del modelo al siguiente periodo. Para ello se utilizaron dos modelos de clasificación distintos; el primero de ellos es Support Vector Machine (SVM) con el objetivo de confeccionar la metodología dinámica, que denominaremos Dynamic Support Vector Machine (D-SVM) y el segundo modelo de clasificación es Linear Penalizad SVM (LP-SVM) con la finalidad de que la metodología construida permita la selección de atributos dinámicamente. Los parámetros utilizados en el modelo de clasificación son; las ventanas de tiempo, ponderadores de relevancia, penalización de los errores y la penalización de los atributos (sólo para el modelo con selección de atributos). De los resultados obtenidos, se utiliza la ventana de tiempo que define el mejor modelo de un periodo y junto a los nuevos datos que se obtengan generan el del próximo.
Esta metodología luego fue aplicada a un caso real en una institución gubernamental chilena (INDAP), en el problema de predicción de comportamiento de pago (credit scoring). Para ello se analizaron 4 instancias de tiempo con 9 atributos para el modelo sin selección de atributos y 18 atributos para el modelo con selección. Luego ambos modelos fueron comparados con uno de clasificación estática, es decir, que las 4 instancias de tiempo son unidas como si fuese una data.
Los resultados obtenidos en esta aplicación son levemente superiores a la metodología estática correspondiente y en el caso de la selección de atributos el modelo utiliza una mayor cantidad.
Las conclusiones de esta investigación son que presenta la ventaja de utilizar una menor cantidad de datos a los disponibles, lo que genera modelos más rápidos y que se van adaptando a los cambios de comportamiento que se producen en el tiempo, al descartar los datos más antiguos en la construcción del nuevo modelo. Con respecto al método con selección de atributos, se destaca que no utiliza un modelo exógeno para seleccionar los atributos sino que el modelo estima los atributos necesarios para cada periodo de tiempo, por lo que se tiene un modelo más estable y generalizado; además se logra obtener información de cómo la relevancia de los atributos cambia en el tiempo. Sobre los resultados se concluye que la metodología D-SVM con y sin selección de atributos es al menos tan buena como los métodos actuales de clasificación.
16
Möller, Manuel. "Spamerkennung mit Support Vector Machines." Thesis, Universitätsbibliothek Chemnitz, 2005. http://nbn-resolving.de/urn:nbn:de:swb:ch1-200500580.
Full textAbstract:
Diese Arbeit zeigt ausgehend von einer Darstellung der theoretischen Grundlagen automatischer Textklassifikation, dass die aus der Statistical Learning Theory stammenden Support Vector Machines geeignet sind, zu einer präziseren Erkennung unerwünschter E-Mail-Werbung beizutragen. In einer Testumgebung mit einem Corpus von 20 000 E-Mails wurden Testläufe verschiedene Parameter der Vorverarbeitung und der Support Vector Machine automatisch evaluiert und grafisch visualisiert. Aufbauend darauf wird eine Erweiterung für die Open-Source-Software SpamAssassin beschrieben, die die vorhandenen Klassifikationsmechanismen um eine Klassifikation per Support Vector Machine erweitert.
17
Liu, Yi. "Studies on support vector machines and applications to video object extraction." Columbus, Ohio : Ohio State University, 2006. http://rave.ohiolink.edu/etdc/view?acc%5Fnum=osu1158588434.
Full text
18
Aleti, Kalyan Reddy. "E-quality control a support vector machines approach /." To access this resource online via ProQuest Dissertations and Theses @ UTEP, 2008. http://0-proquest.umi.com.lib.utep.edu/login?COPT=REJTPTU0YmImSU5UPTAmVkVSPTI=&clientId=2515.
Full text
19
Shilton, Alistair. "Design and training of support vector machines." Connect to thesis, 2006. http://repository.unimelb.edu.au/10187/443.
Full textAbstract:
In this thesis I introduce a new and novel form of SVM known as regression with inequalities, in addition to the standard SVM formulations of binary classification and regression. This extension encompasses both binary classification and regression, reducing the workload when extending the general form; and also provides theoretical insight into the underlying connections between the two formulations.
20
Craddock, Richard Cameron. "Support vector classification analysis of resting state functional connectivity fMRI." Diss., Atlanta, Ga. : Georgia Institute of Technology, 2009. http://hdl.handle.net/1853/31774.
Full textAbstract:
Thesis (Ph.D)--Electrical and Computer Engineering, Georgia Institute of Technology, 2010.Committee Chair: Hu, Xiaoping; Committee Co-Chair: Vachtsevanos, George; Committee Member: Butera, Robert; Committee Member: Gurbaxani, Brian; Committee Member: Mayberg, Helen; Committee Member: Yezzi, Anthony. Part of the SMARTech Electronic Thesis and Dissertation Collection.
21
Saxena, Vishal 1979. "Support vector machine and its applications in information processing." Thesis, Massachusetts Institute of Technology, 2004. http://hdl.handle.net/1721.1/29404.
Full textAbstract:
Thesis (M. Eng.)--Massachusetts Institute of Technology, Dept. of Civil and Environmental Engineering, 2004.Includes bibliographical references (leaves 59-61).
With increasing amounts of data being generated by businesses and researchers there is a need for fast, accurate and robust algorithms for data analysis. Improvements in databases technology, computing performance and artificial intelligence have contributed to the development of intelligent data analysis. The primary aim of data mining is to discover patterns in the data that lead to better understanding of the data generating process and to useful predictions. One recent technique that has been developed to handle the ever-increasing complexity of hidden patterns is the support vector machine. The support vector machine has been developed as robust tool for classification and regression in noisy, complex domains. Current thesis work is aimed to explore the area of support vector machine to see the interesting applications in data analysis, especially from the point of view of information processing.
by Vishal Saxena.
M.Eng.
22
Tuan, Nguyen Thanh, and 阮清俊. "Online Transductive Support Vector Machine." Thesis, 2008. http://ndltd.ncl.edu.tw/handle/24865175611241572360.
Full textAbstract:
碩士大葉大學
電機工程學系
96
The support vector machine (SVM) is one kind of machine learning methods based on the Statistical Learning Theory. The SVM is designed to construct a separating hyperplane between two classes of points, such that the margin between the hyperplane and the points closest to it becomes maximal. However, there are some disadvantages in applying the SVM in pattern classification problems. Firstly, the SVM is usually trained by supervised learning. Thus, the SVM model needs to be retrained from scratch whenever a new sample arrives. Secondly, the SVM trained with only a few labeled data can lead to construct very well performing classification systems, but its generalization ability highly depends on which samples are chosen for training. On the other hand, the labeled data is scarce and expensive to generate while the unlabeled data is often readily available in real world application. To overcome the aforementioned problems, an online trandusctive SVM (OTSVM) is proposed to train the SVM model incrementally with new unlabeled data. The OTSVM is developed to combine the trandusctive SVM model with online learning for classification. Unlike supervised SVM learning, in which no learning occurs when labeling unlabeled samples, the OTSVM can learn from labeled and unlabeled samples progressively. Besides, the OTSVM increases the classification accuracy while keeping the memory requirements and computation complexity at a manageable level. In order to investigate the efficiency and effectiveness of the proposed OTSVM method, examples of linearly/non-linearly separable data and terrain classification of SAR images are carried out to compare with supervised SVM learning, TSVM, PTSVM, and unsupervised learning. From simulation results, we conclude that the OTSVM can maintain acceptable classification accuracy with limited labeled data and large quantity of unlabeled data.
23
Chen, Tsai-Ying, and 陳采瀅. "The Compressed Smooth Support Vector Machine." Thesis, 2016. http://ndltd.ncl.edu.tw/handle/45014192212713031489.
Full textAbstract:
碩士國立臺灣科技大學
資訊工程系
105
In most cases, the real world data is not linearly separable. Under this condition, we can use the nonlinear Support Vector Machines (SVM), which involves nonlinear kernel matrix, instead of the linear SVM as a classification optimization problem. A key problem with that the disadvantages of the nonlinear SVM are apparent when encountering large-scale data in that the size of the kernel matrix grows quadratically with the size of the data which makes in-memory computation intractable. Our aim is to avoid the intractability of the nonlinear SVM on large-scale data. We introduce the Compressed Smooth Support Vector Machine (CSSVM), a new method based on the Smooth SVM (SSVM). Furthermore, we apply the idea of a reduced kernel and combine it with compressed sensing. In our method we use a two stage reduction: the first reduction is logically the same as Reduced SVM (RSVM) but the reduction ratio will not be too small in this stage. That is, we do not throw away nearly as much data as we normally do with standard RSVM. In the second reduction we use a sensing matrix to reduce the dimension. In this way we reduce the dimension without sacrificing valuable information which allows us not to obviously affect the accuracy, while improving stability. The numerical result shows that the CSSVM is competitive in both speed and the stability of accuracy. In conclusion, CSSVM is a good choice for handling large-scale classification problems using a nonlinear separating hyperplane.
24
Hung-Yuan, Tseng. "Face Recognition Using Support Vector Machine." 2005. http://www.cetd.com.tw/ec/thesisdetail.aspx?etdun=U0002-1209200513495300.
Full text
25
Tseng, Hung-Yuan, and 曾宏永. "Face Recognition Using Support Vector Machine." Thesis, 2005. http://ndltd.ncl.edu.tw/handle/84749249050598684400.
Full textAbstract:
碩士淡江大學
航空太空工程學系碩士班
93
Human face detection and recognition plays an important role in application such as video surveillance, personal security and face database management. A novel Support Vector Machines (SVM) is adopted for face recognition. SVM can handle classification problem effectively without establishing the prior knowledge database, and obtain support vector and related margin. To shorten the computing time, a modified version of SVM, namely Lagrangian support vector machine (LSVM) is applied here. An effective method to deal with the eyes and mouth region is proposed in this thesis. We verify the correction rate of the utilize method via a database, CVL, that contains 91 images of 31 individuals.
26
Chien, Li-Jen, and 簡立仁. "Robust Smooth Support Vector Machine Learning." Thesis, 2010. http://ndltd.ncl.edu.tw/handle/81221681928172170409.
Full textAbstract:
博士國立臺灣科技大學
資訊工程系
98
This dissertation proposes four robust smooth support vector machine learning methodologies. First, we propose a new approach to generate representative reduced set for RSVM. Clustering reduced support vector machine (CRSVM) generates cluster centroids of each class and uses them to form the reduced set. By estimating the approximate density for each cluster, we can compute the width parameter used in Gaussian kernel. Secondly, we modify the previous 2-norm soft margin smooth support vector machine (SSVM2) to propose a new 1-norm soft margin smooth support vector machine (SSVM1). We also propose a heuristic method of outlier filtering for SSVMs which costs little in training process and improves the ability of outlier resistance a lot. Thirdly, we introduce the smooth technique into 1-norm SVM and call it smooth LASSO for classification (SLASSO). It can provide simultaneous classification and feature selection. Results showed that SLASSO has slightly better accuracy than other approaches with the desirable ability of feature suppression. In the end of this dissertation, we implement a ternary SSVM (TSSVM) and use it to design a novel multiclass classification scheme, one-vs.-one-vs.-rest (OOR). It decomposes the problem into a series of k(k-1)/2 ternary classification subproblems. Results show that TSSVM/OOR performs better than one-vs.-one and one-vs.-rest. We also find out that the prediction confidence of OOR is significantly higher than the one-vs.-one scheme. Due to the nature of OOR design, it can be applied to detect the hidden (unknown) class directly. We conduct a "leave-one-class-out" experiment on the pendigits dataset which shows that OOR outperforms the one-vs.-one and one-vs.-rest in the hidden class detection rate.
27
Chen, Bo-Juen, and 陳博準. "Two applications of support vector machine." Thesis, 2003. http://ndltd.ncl.edu.tw/handle/72678206990554657768.
Full text
28
Lin, Xiao-Rong, and 林筱榮. "Text Detection Using Support Vector Machine." Thesis, 2008. http://ndltd.ncl.edu.tw/handle/23636792602999015895.
Full textAbstract:
碩士元智大學
資訊工程學系
96
Detecting text from image or video streams can be widely used in a variety of application fields. For example, detecting and extracting text from documents in advance can facilitate following OCR module. For automatic recognition of vehicle license plate, the key point is to locate the position of vehicle license plate successfully. For advertisement design, it is easy to replace text or background if text regions can be automatically extracted. For content-based retrieval, automatic extraction of text regions can lead more efficient and effective retrieval results since text and content can be incorporated to facilitate retrieval process. In this thesis, a new method to detect text using support vector machine (SVM) is proposed. The challenges of this approach include the following two issues. First, feature selection and database construction are two essential processes to achieve optimal SVM. However, it is hard to design the two tasks. Second, texts are often embedded in images and may vary in language, font, size, and deformation, which, in turn enhance the difficulty of the text detection problem. In this thesis, discriminating features are adopted and bootstrap training are involved to construct training database for text detection using SVM. Experimental results prove the effectiveness of the proposed text detection method.
29
Tang, Lung, and 唐龍. "Image Compression using Support Vector Machine." Thesis, 2003. http://ndltd.ncl.edu.tw/handle/53037646816913726174.
Full textAbstract:
碩士義守大學
資訊工程學系
91
Image compression addresses the problem of how to reduce the amount of data required to represent the image and the basic process is the removal of redundant data in the image. In the researches of image compression, there are many kinds of techniques such as vector quantization (VQ), Wavelet transform, fractal theory, well-known standard of compression --- JPEG, GIF and so on. Support vector machine (SVM) is a learning system, which was first introduced by Vapnik in 1992. There are two main categories for support vector machines: support vector classification (SVC) and support vector regression (SVR). SVM can generalize complicated gray level structures with only very few support vectors and thus provides a new mechanism for image compression. In this thesis, the algorithms of SVM are utilized to perform the compression of gray images. The method used is SVR which reduces the number of support vectors of image blocks in order to improve the compression ratio. Meanwhile, the parameters for SVR are selected according to the properties of the block in the frequency domain so as to improve the qualities of the retrieved images.
30
Liao, Ping-Lun, and 廖柄. "Support Vector Machine for Smart Home." Thesis, 2010. http://ndltd.ncl.edu.tw/handle/zkdgmm.
Full textAbstract:
碩士國立虎尾科技大學
資訊工程研究所
98
Smart Home is based on the research of home digitization for human beings. The related researches report that context awareness is able to achieve the applications for a variety of scenarios and perform specific functions, which results in a more comfortable and convenient environment. In addition, the integration system can save energy and lower cost. Therefore, a model for integrating Bluetooth, Zigbee, Universal Serial Bus, WiFi standards is proposed. Those wireless technologies are used for gathering data. To simulate the scenarios of Smart Home, the proposed system substitute CPLDs for home appliances. Machine learning is a theory that enhances a computer system''s knowledge. Support Vector Machines is a machine learning theory that has two major functions that are classification and regression analysis. To enhance the system''s intelligence, Wavelet Support Vector Machine is applied to the system. Besides, the system will be designed by Object-Oriented Design and be analyzed by Object-Oriented Analysis. Therefore, the proposed system will be modeled by Unified Modeling Language. C# is chosen to be the Object-Oriented Language that implements the system software. The experimental results suggest that the proposed model is feasible.
31
Chen, Hsiang-Hsuan, and 陳祥瑄. "Distributed Consensus Reduced Support Vector Machine." Thesis, 2019. http://ndltd.ncl.edu.tw/handle/as23xk.
Full textAbstract:
碩士國立交通大學
應用數學系數學建模與科學計算碩士班
107
Nowadays, Machine learning performs astonishingly in many different fields. The more data we have, our machine learning methods show better results. However, in some cases, the data owners may not want to share the information they have, because those materials contain privacy issues. On the other hand, sometimes we encounter a very large dataset, which are difficult to store in a single machine. To deal with these two problems, we propose the distributed consensus reduced support vector machine (DCRSVM) for binary classification. Imagine that we have many local working units and a central master, and each working unit owns its data. The DCRSVM includes the following two merits. First, our method keeps the privacy of data, so we are not going to disclose local data to the central master. Besides, when we confront a large dataset, which is hard to store in a single server, the central master can still derive a good machine learning model even if the data stores only in local devices. Our method successfully solves the problems we mentioned above, and it generates a competitive result.
32
Das, Anshuman. "Mining Machine Reliability Analysis Using Ensembled Support Vector Machine." Thesis, 2012. http://ethesis.nitrkl.ac.in/3324/1/Anshuman_Das%2C_108MN053.pdf.
Full textAbstract:
Estimation of reliability plays an important role in performance assessment of any system. Reliability predictions are important for various purposes, like production planning, maintenance planning, reliability assessment, fault detection in manufacturing processes, and risk and liability evaluation. In this study, a Support vector machine (SVM)-based model for forecasting reliability was developed. A genetic algorithm was applied for selecting SVM parameters. The developed model was validated by applying two benchmark data sets. A comparative study reveals that the proposed method performs better than existing methods on benchmark data sets. A case study was conducted on a Dumper’s past time-to-failure data, and cumulative time-to-failure was calculated for reliability modeling. The results demonstrate that the developed model performs well with high accuracy (R2 = 0.99) in the failure prediction of a Dumper. These accurate predictions can help a company in making accurate preventive maintenance and accordingly production and equipment planning can help in increasing production.
33
Huang, Zheng-Cheng, and 黃政誠. "Performance analysis of Spark Support Vector Machine." Thesis, 2017. http://ndltd.ncl.edu.tw/handle/aq4ekc.
Full textAbstract:
碩士國立雲林科技大學
資訊管理系
105
Due to the fast-paced nature of technology, the data of network world and real world increase enormously. Nowadays, the technology of computer has been unable gradually to load a huge amount of dataset to accommodate this enormous data, the concept of cloud computing and big data has been developed. The combination of both has brought a lot of commercial transactions. In recent years, the spark in memory through the proposed cluster computing concept, to solve many Hadoop general computation time long problem, because the use spark operation is performed in memory, distributed computing is faster than computing power to solve the disk operation time cost in distributed computing. Therefore this study through the ex-periment environment of traditional classifier experiment environment and spark sin-gle, distributed under the observation of Map dispersion and the number of adjust-ment algorithm performed through experiment, whether the number of view Map distribution have influence on the results, the spark of decentralized is the use of 4 units of the virtual machine environment on Map distribution quantity adjustment. The experimental results show that the more the number of Map distributed compu-ting time is faster depending on the number of parameters and data set size, can also be seen from the experimental results the Map distribution quantity adjustment and virtual machine under the same circumstances, in the computation time can get the advantage.
34
Meng, Chao-Hong, and 孟昭宏. "Phone Recognition using Structural Support Vector Machine." Thesis, 2010. http://ndltd.ncl.edu.tw/handle/45917295138142120792.
Full text
35
Tsai, Ming-Chia, and 蔡明嘉. "Nasal event detection using support vector machine." Thesis, 2010. http://ndltd.ncl.edu.tw/handle/06359129646759103280.
Full text
36
Ying, Chiou Chiun, and 邱群穎. "Monitoring correlated processes using support vector machine." Thesis, 2007. http://ndltd.ncl.edu.tw/handle/07927030419295046584.
Full textAbstract:
碩士輔仁大學
管理學研究所
95
Traditional statistical process control (SPC) techniques are not applicable in many process industries due to the assumption of uncorrelated datasets. Several methods have been proposed to deal with autocorrelated parameters. In this research, we demonstrate that support vector machine (SVM) can be effective tools to identify shifts in process parameter values from AR(1) processes with various values of the autocorrelation coefficient. There are two kinds of monitoring models built in this thesis. One is 2 classifications model identifying if there is any shift existing for generated process data; the other one is 4 classifications model recognizing the quantity of mean shift of the data sets which contain one, two and three standard deviations shifted from non-shifted data and non-shifted data. Theses two models were estimated by average run length (ARL) and correct classification percentages respectively. As the results reveal, SVM performed superiorly than X chart in all conditions in 2 classes experiment; in 4 classes, the proposed SVM model effected greater than time serious control chart in 80% conditions. For , SVM can be above 80% accurate with the 0 shift where time serious control chart is more effective than SVM. Therefore, the results show the possibility that SVM can be applied to improve control in manufacturing processes that generate correlated process data.
37
Lin, Chang-ching, and 林長青. "Support Vector Machine for Discovering Scientific Equations." Thesis, 2003. http://ndltd.ncl.edu.tw/handle/80988014791877451365.
Full textAbstract:
碩士國立雲林科技大學
電子與資訊工程研究所碩士班
91
Support vector machine (SVM) is a popular method of machine learning. Support vector machine has been applied to biotechnology, text categorization and image recognition. Support vector machine has good performance like decision tree and artificial neural network. In this thesis we use SVM to rediscover scientific equations from given scientific data. Finally we compare BACON’s method on discovering equations to support vector machine.
38
Liu, Kun-Xain, and 劉堃憲. "Metagenomic Phylogenetic Classification with Support Vector Machine." Thesis, 2015. http://ndltd.ncl.edu.tw/handle/81513256489784510377.
Full textAbstract:
碩士逢甲大學
資訊工程學系
103
Metagenomics in environmental samples collected each DNA is unidentified DNA may have more sample in environmental. Metagenomics face a major challenge is to analyze the binning of the original sequence to be the same or similar communities in the category. The new sequencing technology makes it easier for metagenomics, simple and quick to get the sequence, but also more difficult to classify, because of short sequences to produce more than the previous technology. Classification short sequence 100 base pairs (bp) have until now relatively inaccurate, researchers need to use the older, long reading technology. We propose Support Vector Machine algorithm with Genetic Algorithm, a classifier to classify metagenome data, trained complete genomes planning and significantly improved the previous classification method based on synthetic.
39
Huang, Chien-Ming, and 黃建銘. "Smile Detection based on Support Vector Machine." Thesis, 2009. http://ndltd.ncl.edu.tw/handle/72875756469700143588.
Full textAbstract:
碩士南台科技大學
資訊工程系
97
The main purpose of this study is to present a method which applies Support Vector Machines which is maintained by two color spaces to Face Detection. First of all, the researcher transferred an image from RGB color space to YCbCr and Normalized Color Coordinates (NCC) color space to reduce the influence which is affected by different light condition to skin sample. After color space translation, the researcher could obtain YCbCr skin and NCC skin, and then found out some skin regions from original image by fuzzy computation. Finally, these skin regions were inputted into Support Vector Machines to distinguish which are or aren't human faces. The result of this research appears that the overall accuracy rate of this system is 97.08%. After face detection, to identify the position of the mouth which is according to eye coordinates, and then the researcher can use curve fitting to calculate the degree of smile. We hope that by applying the result of this research to many situations between users and computer systems that we can get more intuitive interactions between users and systems. For example, robots with this kind of equipment can understand the emotional reaction of human. In others cases, this technique could used to assist doctors to estimate emotion of patients, or for people who are talking by mobile but also understanding the emotion of other people by detecting the emotion of cell phone users, or for some service business to remind their staff to always maintain a smile to attract more customers.
40
Huang, Bo-Jhen, and 黃柏禎. "Radar Precipitation Estimation Using Support Vector Machine." Thesis, 2015. http://ndltd.ncl.edu.tw/handle/60276377877488607102.
Full textAbstract:
碩士臺北市立大學
資訊科學系碩士在職專班
103
Significant problems exist for traditional precipitation estimates from radar when it comes to comparing with rain gauges on the ground. To avoid the problem and better utilize the information in the four-dimensional structure of the atmosphere, this research proposes a precipitation estimation method based on support vector machine and regression in the hopes to improve the accuracy of precipitation estimates. To generate various feature sets as inputs for the experiments, a set of tools have been developed, including a radar feature extraction and labeling tool based on AWIPS II (Advanced Weather Interactive Processing System) and a library for further batched processing of the extracted features such as calculation of statistics and combinations of features from various elevations and/or products. For experiments of this research, Taiwan Central Weather Bureau’s Wu-Fen-Shan weather radar products, including reflectivity (Z), differential reflectivity (ZDR) and specific differential phase (KDP) from rainy days in the years of 2012, 2013 and 2014 are used as the source of features, and the target location for precipitation estimation is Taipei weather station. In one of the experiments, we first aggregated over the feature vectors within the same observation hour to derive a mean feature vector, then partitioned 10-minute precipitation accumulations into 6 classes as labels of the features, by using the resulting feature set along with support vector regression, the best result gives a root mean squared error of 0.54 and correlation coefficient of 0.95, corresponding to features derived from 0.5-degree elevation specific differential phase (KDP), when the extreme 10-minute precipitation value of 9.4(mm) in the test set is removed, the corresponding root mean squared error becomes 0.5 and correlation coefficient dropped to 0.72. Results from this reaearch also indicate the proposed support vector regression estimation method has about the same performance as traditional R(KDP) given by Sachidananda and Zrnic(1987) and is better than R(ZDR,KDP) by Ryzhkov and Zrnic(1995). In summary, using combinations of statistics such as middle value, mean, and maximum 5 values calculated from 5x5 feature vectors of low-elevation KDP products, along with support vector regression, is the best solution found for radar precipitation estimation in this research. While combinations of Z、ZDR and KDP as well as combinations of various elevations are also tried, no significant improvements can be derived in this research.
41
Tsai, Yu Shiue, and 蔡豫學. "Reservoir Drought Prediction Using Support Vector Machine." Thesis, 2011. http://ndltd.ncl.edu.tw/handle/32942032852924958649.
Full textAbstract:
碩士國立屏東科技大學
水土保持系所
99
Drought often occurs without signs. When facing drought disasters, the best moment for making water restriction policy has already passed. The build of early warning system of drought can help better allocate limited water resource. In this study, five models (Support Vector Machine, SVM, Artificial Neural Network, ANN, multi-variables regression, maximum likelihood and Baysin classifier) were established to predict reservoir drought in next 1-9 ten-days in Tsengwen Reservoir.The discrimination drought prediction models use three ways; two class- drought conditions (distinguish whether the drought), two class- severe drought conditions (distinguish whether the severe drought), the three class (non-drought, drought, severe drought),to estimate the next 1-9 ten-days in Tsengwen Reservoir of drought phenomenon, and assessment of the model in different circumstances to determine the performance of drought prediction. Expect to define the specific stage of drought conditions in order to release the drought warning or implement drought emergency measures. The results show that severe drought conditions and drought conditions in SVM and ANN has good performance and in these two conditions, the differentiating of severe-drought and drought category is SVM better than ANN. The prediction accuracy of SVM is 5-10% higher than ANN. The prediction accuracy of non-drought and severe drought categories of SVM in the next 1-9 ten-days are significantly better than other prediction models, the performance of ANN prediction in drought category is better than other prediction models. To improve the prediction results, in this study two stage classification of support vector machine model is used to predict of reservoir drought. The results show that the predicted accuracy of two stage classification (in drought category) is 15-30% higher than the three classifications.
42
Su, Yen-Ting, and 蘇彥庭. "TAIEX Trend Prediction with Support Vector Machine." Thesis, 2016. http://ndltd.ncl.edu.tw/handle/61892041901017770533.
Full textAbstract:
碩士國立中山大學
財務管理學系研究所
104
This paper develops a prediction model which combined genetic algorithms and support vector machine to predict the five-day-ahead direction of the TAIEX. We employ technical indicators as model input features, and the model output is upward or downward signal. To confirm the model output signal is able to make return from market, we use the output signals to do backtesting on TAIFEX between 2000 and 2014. In addition, we also try to use clustering and feature selection method to training data for improving the accuracy rate and investment return. The best prediction model is used clustering and no feature selection method to the training data. The accuracy rate of all sample is 52.66%, and the accuracy rate of the sample on the condition of predicting upward signal is 57.15%, but the accuracy rate of the sample on the condition of predicting downward signal is only 48.29%. Because our models have the higher accuracy rate on the condition of predicting upward signal, we only use upward signal to do backtesting on TAIFEX. In the backtesting, we consider the impact of stop loss quantile, stop loss rate, and leverages on the investment return. The best backtesting result of the model is used clustering and F-score feature selection method to the training data, its annualized rate of return can reach 23.47% when the the stop loss quantile sets to 0.5, stop loss rate sets to 10% and leverage sets to 3 times. This study demonstrated that machine learning model can be used in securities investment. It can obtain abundant reward from market through our method.
43
Lin, Guan-Jhong, and 林冠中. "Sequential Support Vector Machine for Face Recognition." Thesis, 2005. http://ndltd.ncl.edu.tw/handle/47391752650312695629.
Full textAbstract:
碩士國立成功大學
資訊工程學系碩博士班
93
One effective approach is to adapt face models using new data in new environment. This thesis presents a novel sequential learning algorithm for the support vector machine（SVM） based face recognition system. First of all, we use a aussian probability model to represent the randomness of SVM parameters. The recursive Bayes theory is applied to sequentially update a posteriori distribution of SVM parameters. The estimated mean vector is adopted to build the output distribution of SVM. During test phase, we classify a test image according to output distributions of two SVM classes. In this study, we demonstrate that the proposed sequential SVM can meet the standard properties of SVM, or equivalently, minimization of classification errors and maximization of distance of output distributions of two classes. In the experiments on using ORL and FERET facial databases, the proposed sequential SVM did improve face recognition accuracy when increasingly enrolling new face adaptation data.
44
Wang, Shiang-Ming, and 王祥銘. "Protein Crystallization Prediction Using Support Vector Machine." Thesis, 2008. http://ndltd.ncl.edu.tw/handle/08623401823795019500.
Full textAbstract:
碩士淡江大學
資訊工程學系碩士班
96
In structural genomics, proteins are essential materials that define life. A protein’s function is strongly related to it’s structure. The ultimate goal of structural research is to determine the three-dimensional structure of a protein. However, structure determination is often a time-consuming and expensive process. Also the process of experimental determination of protein structure has a high ratio of failures at different stages. There are two prevalent methods for protein structure determination - the magnetic resonance (NMR) spectroscopy and X-ray crystallography. The NMR protein structure determination requires weeks of data acquisition, expensive stable isotope labeling, and extensive manual analysis of data. On the other hand, X-ray crystallography is by far the most successful approach to structure determination. But X-ray crystallography has an importance condition. That is the protein target must be crystallized first. Then the resulting crystal must diffract to sufficient resolution. Therefore, prediction of protein crystallization is an essential problem for structural research. Protein Data Bank (PDB) provides us detailed protein sequence information. We use information from a protein’s primary structure, i.e. the amino acid sequence, as the input to the support vector machine to predict the protein’s crystallizability. Several protein features that correlate with protein crystallization are identified first. The support vector machine then generates a hyperplane in the feature space to predict the protein sequence’s crystallizability. We also investigated two feature selection methods - the wrapper method and the filter method. The purpose is to remove irrelevant and redundant features, and thus reduce dimensionality of the input data. A feature subset can be resulted to make the support vector machine result in higher prediction accuracy. The feature selection approach can also help us recognize which protein features are more important for protein crystallization. This can help the chemist to understand the key factors of protein crystallization. An overall prediction accuracy of 79% was achieved on a screened PDB data set with 5-fold cross-validation. The true-positive rate (crystallization) is 80.8% and the true-negative (non-crystllizable) rate is 78.4%.
45
Yang, Jyun-Siang, and 楊竣翔. "Human Face Identical using Support Vector Machine." Thesis, 2017. http://ndltd.ncl.edu.tw/handle/5t5894.
Full textAbstract:
碩士國立勤益科技大學
電子工程系
105
In this paper, the human face recognition is used for the purpose of research. The study takes two corners and the center point of eyes in the face, and two corners and the center point of the mouth as the feature points. The six feature points connected to each other are used as the eigenvector, and, finally, the face image recognition is completed by support vector machine technology combined with the eigenvector. In the course of the experiment, the face’s upper and lower parts respectively adapt skin color classification and texture technology, and use the histogram technique to find the ROI area of the eye and mouth. In the ROI area, Sobel Edge Detection is used to find two corners of eyes and two corners of the mouth (a total of four feature points),, and then the center point of the two corners separately (a total of 2 feature points). These six feature points are used to calculate the six eigenvectors as the input parameters of the support vector machine. Finally, the face vector recognition is successfully performed by the support vector machine. It is proved that the method can correctly identify the target face in different face data.
46
Tsai, Liang-yu, and 蔡亮宇. "GPS Multipath Mitigation Using Support Vector Machine." Thesis, 2018. http://ndltd.ncl.edu.tw/handle/j9e75t.
Full textAbstract:
碩士國立臺灣海洋大學
通訊與導航工程學系
106
The positioning accuracy of GNSS is generally great in an open-sky environment. However, it usually exhibits poor positioning results in urban canyon environments due to pseudo-range measurement errors caused by multipath effect, which leads to performance degradation of the entire positioning system. For this reason, in order to maintain the accuracy of positioning systems, it is necessary to determine when the GNSS positioning is accurate and which satellites can have their pseudo-range measured accurately without multipath errors. Thus, the objective of our work is to detect the multipath errors in the satellite signals, estimate pseudo-range and line of sight (LOS) probability, and use support vector machine (SVM) to weight the satellites and exclude these signals to improve the positioning accuracy of GNSS.
47
Amayri, Ola. "On email spam filtering using support vector machine." Thesis, 2009. http://spectrum.library.concordia.ca/976212/1/MR63317.pdf.
Full textAbstract:
Electronic mail is a major revolution taking place over traditional communication systems due to its convenient, economical, fast, and easy to use nature. A major bottleneck in electronic communications is the enormous dissemination of unwanted, harmful emails known as "spam emails". A major concern is the developing of suitable filters that can adequately capture those emails and achieve high performance rate. Machine learning (ML) researchers have developed many approaches in order to tackle this problem. Within the context of machine learning, support vector machines (SVM) have made a large contribution to the development of spam email filtering. Based on SVM, different schemes have been proposed through text classification approaches (TC). A crucial problem when using SVM is the choice of kernels as they directly affect the separation of emails in the feature space. We investigate the use of several distance-based kernels to specify spam filtering behaviors using SVM. However, most of used kernels concern continuous data, and neglect the structure of the text. In contrast to classical blind kernels, we propose the use of various string kernels for spam filtering. We show how effectively string kernels suit spam filtering problem. On the other hand, data preprocessing is a vital part of text classification where the objective is to generate feature vectors usable by SVM kernels. We detail a feature mapping variant in TC that yields improved performance for the standard SVM in filtering task. Furthermore, we propose an online active framework for spam filtering. We present empirical results from an extensive study of online, transductive, and online active methods for classifying spam emails in real time. We show that active online method using string kernels achieves higher precision and recall rates.
48
Chen, Po-Hao, and 陳伯豪. "Stock Indices Forecasting Using a Support Vector Machine." Thesis, 2004. http://ndltd.ncl.edu.tw/handle/89862312651459498965.
Full textAbstract:
碩士國立交通大學
科技管理研究所
92
This thesis deals with the application of a novel neural network technique, Support Vector Machine (SVM), in stock indices movement prediction. The purpose of this thesis is to demonstrate and verify the predictability of stock index direction using SVM, to develop effective trading strategies and to test the relative performance. A real future contract (Taiwan Stock Exchange Capitalization Weighted Stock Index) collected from Taiwan Futures Exchange is used as the data set. The series of relative difference in percentage of price (RDP) is adopted as the input variables to describe the patters of market movement. Results indicate that the technique is capable of returning results that are superior to those attained by buy-and-hold strategy.
49
Chang, Chih-Cheng, and 張志丞. "Smooth Support Vector Machine for Multi-class Classification." Thesis, 2007. http://ndltd.ncl.edu.tw/handle/n7sreh.
Full textAbstract:
碩士國立臺灣科技大學
資訊工程系
95
The smooth technique has been introduced into SVMs for both binary classification and regression problems. In this work, we extend the previous work in Smooth Support Vector Machine (SSVM) from binary to k-class classification based on a single machine approach and call it Multi-class Smooth Support Vector Machine (MSSVM). Similar to other single machine approaches, MSSVM will become very complicated and very difficult to solve when k is large. We only implement MSSVM for trinary classification problems. For k>3, we propose a Smooth One-vs.-One-vs.-Rest (SOOR) scheme which decomposes a k-class classification problem into k(k-1)/2 trinary classification problems. Each trinary classification problem is defined by two particular classes and the rest part of data. The class label is determined by a majority voting scheme. Under SOOR setting, we can filter out those nuisance votes in One-vs.-One scheme. In our experiments, we compare MSSVM and SOOR with One-vs.-One and One-vs.-Rest, which are simple and well-known schemes for multi-class classification, on nine public UCI data sets. The results show that MSSVM and SOOR have a slightly better accuracy than One-vs.-One and One-vs.-Rest schemes, and the confidence of SOOR is obviously higher than One-vs.-One for prediction.
50
黃泓偉. "To identify nonobjective images by Support Vector Machine." Thesis, 2017. http://ndltd.ncl.edu.tw/handle/tu9d2j.
Full textAbstract:
碩士國立高雄應用科技大學
資訊工程系
105
One of the difficulties in searching natural images is the semantic gap between low-level pixel data and the content that the image is perceived by human. Most of the research focus on entity images which often contain objects in it, such as dinosaurs, flowers and buses. On the contrary, to search for nonobjective images, for example, art, emotion, friendship, time and freeze, is a much more difficult challenge than typical entity images retrieval. Because nonobjective images can only be judged by the surrounding environment, feelings or subjective understandings and they usually do not have clear or consistent objects in it. Thus, we first manually classified selected images into the five categories aforementioned. Then we built SVMs to train and test on these pre-classified nonobjective images. Finally, the results were compared with common retrieval methods. We hope that our approach and results of this thesis will contribute to the identification and analysis of nonobjective images, and bring the semantic gap closer in the future.