Harris Drucker

Monmouth University · Department of Software Engineering

Three-screen services provide the right solution for consumers to access rich multimedia resources by any device, anytime and anywhere. In this paper, we describe a prototype system of content personalization and adaptation for three-screen services. The system continuously acquires content from TV broadcast feeds, indexes and adapts the content fo...

IPTV customers will have access to thousands of video content sources and will require powerful yet intuitive tools to locate desired content. We propose a solution based on stored user interest profiles and multimodal processing for content segmentation to produce manageable content subsets for users. Segmentation involves part-of-speech tagging o...

This paper introduces a methodology for speech data mining along with the tools that the methodology requires. We show how they increase the productivity of the analyst who seeks relationships among the contents of multiple utterances and ultimately must link some newly discovered context into testable hypotheses about new information. While, in it...

This paper introduces a methodology for speech data mining along with the tools that the methodology requires. We show how they increase the productivity of the analyst who seeks relationships among the contents of multiple utterances and ultimately must link some newly discovered context into testable hypotheses about new information. While in its...

Contents 0 A Roadmap 6 0.1 How to read this Thesis . . . . . . . . . . . . . . . . . . . . . . . 6 0.2 A Short Review of Approximation and Regression Estimation . . 7 0.3 The Reason for Support Vectors . . . . . . . . . . . . . . . . . . . 8 1 Introduction 10 1.1 The Regression Problem . . . . . . . . . . . . . . . . . . . . . . . 10 1.2 A Special...

A new regression technique based on concept of support vectors is introduced. We compare support vector regression with a committee regression technique (bagging) based on regression trees and ridge regression done in feature space. On the basis of these experiments, it is expected that SVR will have advantages in high dimensionality space because...

We compare support vector machines (SVMs) to Rocchio, Ide regular and Ide dec-hi algorithms in information retrieval (IR) of text documents using relevancy feedback. It is assumed a preliminary search finds a set of documents that the user marks as relevant or not and then feedback iterations commence. Particular attention is paid to IR searches wh...

Generating an architecture for an ensemble of boosting machines involves making a series of design decisions. One design decision is whether to use simple “weak learners” such as decision tree stumps or more complicated weak learners such as large decision trees or neural networks. Another design decision is the training algorithm for the constitue...

We show that support vectors machines

We show that support vectors machines (SVM's) are much better than conventional algorithms in a relevancy feedback (RF) environment in information retrieval (IR) of text documents. We track performance as a function of feedback iteration and show that while the conventional algorithms do very well in the initial feedback iteration if the topic sear...

This paper compares the performance of several classifier algorithms on a standard database of handwritten digits. We consider not only raw accuracy, but also training time, recognition time, and memory requirements. When available, we report measurements of the fraction of patterns that must be rejected so that the remaining patterns have misclass...

This paper compares the performance of several classifier algorithms on a standard database of handwritten digits. We consider not only raw accuracy, but also training time, recognition time, and memory requirements. When available, we report measurements of the fraction of patterns that must be rejected so that the remaining patterns have misclass...

This paper compares the performance of several classi#er algorithms on a standard database of handwritten digits. We consider not only raw accuracy, but also rejection, training time, recognition time, and memory requirements. 1 COMPARISON OF LEARNING ALGORITHMS FOR HANDWRITTEN DIGIT RECOGNITION Y. LeCun, L. Jackel, L. Bottou, A. Brunot, C. Cortes,...

We study the use of support vector machines (SVM) in classifying e-mail as spam or nonspam by comparing it to three other classification algorithms: Ripper, Rocchio, and boosting decision trees. These four algorithms were tested on two different data sets: one data set where the number of features were constrained to the 1000 best features and anot...

A splitting criterion that arrives out of the context of a new boosting algorithm is used to construct classification trees. Trees constructed using this Z function are compared to those using the entropy function of C4.5 and are found to give much lower error rates. The Z function is also used to construct boosting machines which, when compared to...

Boosting is a method to construct a committee of weak learners that lowers the error rate in classification and prediction error in regression. Boosting works by iteratively constructing weak learners whose training set is conditioned on the performance of the previous members of the ensemble. In classification, we train neural networks using stoch...

A new regression technique based on Vapnik's concept of support vectors is introduced. We compare support vector regression (SVR) with a committee regression technique (bagging) based on regression trees and ridge regression done in feature space. On the basis of these experiments, it is expected that SVR will have advantages in high dimensionality...

In the regression context, boosting and bagging are techniques to build a committee of regressors that may be superior to a single regressor. We use regression trees as fundamental building blocks in bagging committee machines and boosting committee machines. Performance is analyzed on three non-linear functions and the Boston housing database. In...

A new regression technique based on Vapnik's concept of support vectors is introduced. We compare support vector regression (SVR) with a committee regression technique (bagging) based on regression trees and ridge regression done in feature space. On the basis of these experiments, it is expected that SVR will have advantages in high dimensionality...

A new boosting algorithm of Freund and Schapire is used to improve the performance of an ensemble of decision trees which are constructed using the information ratio criterion of Quinlan's C4.5 algorithm. This boosting algorithm iteratively constructs a series of decision trees, each decision tree being trained and pruned on examples that have been...

The digital battlefield will present unprecedented requirements for the transfer of digital information (voice, data, and imagery). Realizing the vision of the Army's 21st century information transport architecture will require application of advanced modeling and simulation technology for performing architecture analysis, “what-if” drills, systems...

This paper compares the performance of classifier algorithmson a standard database of handwritten digits. We consider not rawaccuracy, but rejection, training time, recognition time, and memoryrequirements."Comparison of Leaning for Handwritten Digit Recognition", International Conference onNeural F. and P. Cie Publishers, 1995Y. Le L. Bottou, C. J...

A new boosting algorithm of Freund and Schapire is used to improve the performance of decision trees which are constructed usin: the information ratio criterion of Quinlan's C4.5 algorithm. This boosting algorithm iteratively constructs a series of decision tress, each decision tree being trained and pruned on examples that have been filtered by pr...

In an optical character recognition problem, we compare (as a function of training set size) the performance of three neural network based ensemble methods (two versions of boosting and a committee of neural networks trained independently) to that of a single network. In boosting, the number of patterns actually used for training is a subset of all...

This paper compares the performance of several classifier algorithms on a standard database of handwritten digits. We consider not only raw accuracy, but also training time, recognition time, and memory requirements. When available, we report measurements of the fraction of patterns that must be rejected so that the remaining patterns have misclass...

We compare the performance of three types of neural network-based ensemble techniques to that of a single neural network. The ensemble algorithms are two versions of boosting and committees of neural networks trained independently. For each of the four algorithms, we experimentally determine the test and training error curves in an optical characte...

A boosting algorithm, based on the probably approximately correct (PAC) learning model is used to construct an ensemble of neural networks that significantly improves performance (compared to a single network) in optical character recognition (OCR) problems. The effect of boosting is reported on four handwritten image databases consisting of 12000...

In order to generalize from a training set to a test set, it is desirable that small changes in the input space of a pattern do not change the output components. This can be done by forcing this behavior as part of the training algorithm. This is done in double backpropagation by forming an energy function that is the sum of the normal energy term...

One test of a new training algorithm is how well the algorithm generalizes from the training data to the test data. A new neural net training algorithm termed double backpropagation improves generalization in character recognition by minimizing the change in the output due to small changes in the input. This is accomplished by minimizing the normal...

One test of a training algorithm is how well the algorithm generalizes from the training data to the test data. It is shown that a training algorithm termed double back-propagation improves generalization by simultaneously minimizing the normal energy term found in back-propagation and an additional energy term that is related to the sum of the squ...

The author shows how to implement a neural net expert system that is optimum in the minimum error sense and recognizes objects based on feature extraction. The expert system can handle features that may not be appropriate to describe certain objects (termed don't care features) or features that cannot be extracted (unknowns or don't know features)...

The implementation of a neural network acting as an expert system is discussed. In some cases a particular feature may not be appropriate to describe a particular object (don't care condition) or a particular feature cannot be extracted or is missing (don't know feature). It is shown how the traditional feedforward neural network cannot be trained...

Monmouth College instituted a Master’s degree program in Software Engineering in the fall of 1985 as a result of extensive collaboration with high technology industries and encouragement from the state for more interaction between academia and industry. We define software engineering as the technological and managerial discipline concerned with sys...

The authors investigated the effect of slips on two facsimile machines and found that a slip can cause a maximum of eight missing lines. At the maximum this creates a missing 0.08 inches of vertical space using a vertical resolution of approximately 100 lines per inch. The receiving facsimile machine does not report these errors to the user. The au...

The PLRS/JTIDS Hybrid (PJH) network is a position/location reporting and communication network composed of Position Location Reporting System (PLRS) User Units, Joint Tactical Information Distribution System (JTIDS) terminals and a net control station (NCS). A study is made of the advisability of using the commercially oriented X.25 protocol for ho...

First Page of the Article

A two-band compression system was designed to improve the intelligibility of speech in noise for persons suffering from sensori-neural hearing impairment. Experiments were carried out at 0 and +6 dB signal-to-noise (S/N) ratio with expansion below compression threshold. Six subjects had mean discrimination scores at S/N of 6 dB of 46, 77, and 87 pe...

This study gives preliminary results in the real-time recognition of ten vowel-like sounds in a continuous speech environment for 20 speakers. The objective is to implement networks with a minimum sum of false alarm and miss rates. A procedure to optimize recognition networks using a speech data base and a computer to vary parameters of recognition...

One of the problems in the design of continuous speech recognition systems is evaluation of the stage which detects occurrences of phonemes in the speech signal. A method of evaluating and optimizing phoneme recognition networks has been developed which uses data obtained by manual 'marking' of phoneme locations on a tape recording of speech. With...

This paper discusses optimization and implementation of recognition networks using interconnections of a standard network element to form a classification network. The standard element has a nonlinear transfer function whose inputs may be weighted by selected resistors. It is assumed that a training set of samples to be accepted or rejected is avai...

A method of separating noiselike sounds from vowel/vowellike sounds where the speech is of telephone quality is presented. This procedure involves the use of the Ïǂ2 nonparametric test, which analyzes the fluctuations of the time waveform and separates the speech into two classes depending on the randomness of the waveform. Initial computer simulat...

Classical communication theory considers preprocessing of speech for more efficient transmission through a noisy channel. However, when a high ambient noise environment is present prior to modulation, conventional techniques break down. This paper discusses speech processing in a high ambient noise environment and how the intelligibility can be imp...