Evaluation of a set of new ORF kernel functions of SVM for speech recognition

Zhang, Xueying; Liu, Xiaofeng; Wang, Zizhong John

doi:10.1016/j.engappai.2013.04.008

Cited by 23 publications

(10 citation statements)

References 16 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…It is also worth noting, that MELC parameter has much more clear geometrical interpretation than SVM's parameter C (which can be seen either as abstract weight of training errors or as an upper bound on the size of Lagrange multipliers). The parameter γ, or in general the formula for V A (v), gives the estimation of optimal kernel width in one dimensional projection of A on v. There are many existing studies [29,30,31] and formulas for such objects, in particular it is possible to perform adaptive kernel width [32] where each point x have its own kernel width σ x .…”

Section: Uci Binary Classificationmentioning

confidence: 99%

Multithreshold Entropy Linear Classifier: Theory and applications

Czarnecki

Tabor

2015

Expert Systems with Applications

View full text Add to dashboard Cite

Linear classifiers separate the data with a hyperplane. In this paper we focus on the novel method of construction of multithreshold linear classifier, which separates the data with multiple parallel hyperplanes. Proposed model is based on the information theory concepts -namely Renyi's quadratic entropy and Cauchy-Schwarz divergence.We begin with some general properties, including data scale invariance. Then we prove that our method is a multithreshold large margin classifier, which shows the analogy to the SVM, while in the same time works with much broader class of hypotheses. What is also interesting, proposed method is aimed at the maximization of the balanced quality measure (such as Matthew's Correlation Coefficient) as opposed to very common maximization of the accuracy. This feature comes directly from the optimization problem statement and is further confirmed by the experiments on the UCI datasets.It appears, that our Entropy Multithreshold Linear Classifier (MELC) obtaines similar or higher scores than the ones given by SVM on both synthetic and real data. We show how proposed approach can be benefitial for the cheminformatics in the task of ligands activity prediction, where despite better classification results, MELC gives some additional insight into the data structure (classes of underrepresented chemical compunds).

show abstract

Section: Uci Binary Classificationmentioning

confidence: 99%

Multithreshold Entropy Linear Classifier: Theory and applications

Czarnecki

Tabor

2015

Expert Systems with Applications

View full text Add to dashboard Cite

show abstract

“…Selecting features [8] utilizing ABC and Nearest Neighbor were introduced for steganalysis in images. Support Vector Machine (SVM) stood prevailing in sorting compared to -NN classifier and ANN by reason of its exceptional classification accuracy and generalization act [9] [10].…”

Section: Literature Reviewmentioning

confidence: 99%

Hybrid Classification Technique for Accurate Detection of DDoS Attacks

Vani. V*,

2020

IJITEE

View full text Add to dashboard Cite

Intellectual intrusion detection system can merely be build if there is accessibility to an effectual data set. A high dimensional quality dataset that imitates the real time traffic facilitates training and testing an intrusion detection system. Since it is complex to scrutinize and extort knowledge from high-dimensional data, it is identified that feature selection is a preprocessing phase during attack defense. It increases the classification accuracy and reduces computational complexity by extracting important features from original data. Optimization schemes can be utilized on the dataset for selecting the features to find the appropriate subspace of features while preserving ample accuracy rate characterized by the inventive feature set. This paper aims at implementing the hybrid algorithm, ABC-LVQ. The bio-inspired algorithm, Artificial Bee Colony (ABC) is adapted to lessen the amount of features to build a dataset on which a supervised classification algorithm, Linear Vector Quantization (LVQ) is applied, thus achieving highest classification accuracy compared to k-NN and LVQ. The NSL-KDD dataset is scrutinized to learn the efficiency of the proposed algorithm in identifying the abnormalities in traffic samples within a specific network.

show abstract

“…When all employed bees complete the search for new food sources, the fitness values of new food sources are calculated and compared to the old ones according to the greedy selection mechanism of The operation procedure of ABC algorithm --Initialization phase --(1) Initialize the population of solutions and assign the population to employed bees (2) while (cycle = MAXcycles) do --Employed bee phase -- (3) for = 1 to SN do (4) Produce a new solution V for employed bees and calculate its fitness value (5) Apply the greedy selection mechanism between V and , select the better one (6) If the solution does not update, the non-updated number trial = trial + 1; otherwise trial = 0 (7) end for --Onlooker bee phase -- (8) Calculate the selection probability (9) = 0, = 1 (10) while ( < ) do (11) if random < then (12) = + 1 (13) Produce a new solution V for onlooker bee of the solution and calculate its fitness value (14) Apply the greedy selection mechanism between V and , select the better one (15) If the solution does not update, trial = trial + 1; otherwise trial = 0 (16) end if (17) = + 1 (18) If = + 1, = 1 (19) end while ( = ) --Scout bee phase -- (20) if trial > LIMIT then (21) Replace with a new random solution V (22) end if (23) Memorize the current optimal solution (24) cycle = cycle + 1 (25) end while (cycle = MAXcycles) Algorithm 1: The pseudocode of ABC algorithm. the amounts and the positions of food sources with onlooker bees.…”

Section: Artificial Colony Bee (Abc)mentioning

confidence: 99%

“…In [14], a feature selection technique based on the ABC and -Nearest Neighbor ( -NN) was employed for image steganalysis. Compared with ANN and -NN classifier, support vector machine (SVM) is a more powerful classification method due to its excellent classification accuracy and generalization performance [15,16].…”

Section: Introductionmentioning

confidence: 99%

An Enhanced Artificial Bee Colony-Based Support Vector Machine for Image-Based Fault Detection

Chen

Zhang

Wang

et al. 2015

Mathematical Problems in Engineering

Self Cite

View full text Add to dashboard Cite

Fault detection has become extremely important in industrial production so that numerous potential losses caused from equipment failures could be saved. As a noncontact method, machine vision can satisfy the needs of real-time fault monitoring. However, image-based fault features often have the characteristics of high-dimensionality and redundant correlation. To optimize feature subsets and SVM parameters, this paper presents an enhanced artificial bee colony-based support vector machine (EABC-SVM) approach. The method is applied to the image-based fault detection for the conveyor belt. To improve the optimized capability of original ABC, the EABC algorithm introduces two enhanced strategies including the Cat chaotic mapping initialization and current optimum based search equations. Several UCI datasets have been used to evaluate the performance of EABC-SVM and the experimental results show that this approach has better classification accuracy and convergence performance than the ABC-SVM and other ABC variants-based SVM. Furthermore, the EABC-SVM can achieve a significant detection accuracy of 95% and reduce the amount of features about 65% in the conveyor belt fault detection.

show abstract

Evaluation of a set of new ORF kernel functions of SVM for speech recognition

Cited by 23 publications

References 16 publications

Multithreshold Entropy Linear Classifier: Theory and applications

Multithreshold Entropy Linear Classifier: Theory and applications

Hybrid Classification Technique for Accurate Detection of DDoS Attacks

An Enhanced Artificial Bee Colony-Based Support Vector Machine for Image-Based Fault Detection

Contact Info

Product

Resources

About