A Data Scientist typically performs a number of tedious and time-consuming steps to derive insight from a raw data set. The process usually starts with data ingestion, cleaning, and transformation (e.g. outlier removal, missing value imputation), then proceeds to model building, and finally a presentation of predictions that align with the end-users objectives and preferences. It is a long, complex, and sometimes artful process requiring substantial time and effort, especially because of the combinatorial explosion in choices of algorithms (and platforms), their parameters, and their compositions. Tools that can help automate steps in this process have the potential to accelerate the time-to-delivery of useful results, expand the reach of data science to non-experts, and offer a more systematic exploration of the available options. This work presents a step towards this goal.

Road traffic sensors provide rich multivariable datastreams about the current traffic conditions. Occasionally, there are unusual traffic events (such as accidents, jams, and severe weather) that disrupt the expected road traffic conditions. Detecting the occurrence of such events in an online and real-time manner is useful to drivers in planning their routes and in the management of the transportation infrastructure. We propose a new method for detecting traffic events that impact road traffic conditions by extending the Bayesian robust principal component analysis (RPCA) approach. Our method couples multiple traffic datastreams so that they share a certain sparse structure. This sparse structure is used to localize traffic events in space and time. The traffic datastreams are measurements of different physical quantities (e.g., traffic flow and road occupancy) by different nearby sensors. Our proposed method processes datastreams in an incremental way with small computational cost; hence, it is suitable to detect events in an online and real-time manner. We experimentally analyze the detection performance of the proposed coupled Bayesian RPCA (BRPCA) using real data from loop detectors on the Minnesota I-494. We find that our method significantly improves the detection accuracy when compared with the traditional PCA and noncoupled BRPCA.

Bayesian reasoning provides an ideal basis for representing and manipulating uncertain knowledge, with the result that many interesting algorithms in machine learning are based on Bayesian inference. In this paper, we use the Bayesian approach with one and two levels of inference to model the semisupervised learning problem and give its application to the successful kernel classifier support vector machine (SVM) and its variant least-squares SVM (LS-SVM). Taking advantage of Bayesian interpretation of LS-SVM, we develop a semisupervised learning algorithm for Bayesian LS-SVM using our approach based on two levels of inference. Experimental results on both artificial and real pattern recognition problems show the utility of our method.

The emergence of large-field radio telescopes has generated the need to process huge amounts of data “on the fly” in order to avoid substantial costs in storage and signal processing. In this paper, we describe a streaming software approach to radio astronomy, and in particular, show how the IBM-developed stream computing architecture System S enables seamless “on the fly” radio astronomy imaging. This approach overcomes pertinent issues of memory limitation and latencies related to data processing, and it allows autonomous processor allocation to tasks as the need arises.

With the widespread adoption of location tracking technologies like GPS, the domain of intelligent transportation services has seen growing interest in the last few years. Services in this domain make use of real-time location-based data from a variety of sources, combine this data with static location-based data such as maps and points of interest databases, and provide useful information to end-users. Some of the major challenges in this domain include i) scalability, in terms of processing large volumes of real-time and static data; ii) extensibility, in terms of being able to add new kinds of analyses on the data rapidly, and iii) user interaction, in terms of being able to support different kinds of one-time and continuous queries from the end-user. In this paper, we demonstrate the use of IBM InfoSphere Streams, a scalable stream processing platform, for tackling these challenges. We describe a prototype system that generates dynamic, multi-faceted views of transportation information for the city of Stockholm, using real vehicle GPS and road-network data. The system also continuously derives current traffic statistics, and provides useful value-added information such as shortest-time routes from real-time observed and inferred traffic conditions. Our performance experiments illustrate the scalability of the system. For instance, our system can process over 120000 incoming GPS points per second, combine it with a map containing over 600,000 links, continuously generate different kinds of traffic statistics and answer user queries.

We consider the problem of estimating the direction of arrival (DOA) of multiple waves incident on a single tripole sensor. Using the physical properties of the electric and magnetic fields, we show that we can disambiguate the DOA of multiple simultaneously incident waves using a set of time sampled 3D measurements from the single sensor. This is different from the traditional approach that uses arrays of antennas to estimate DOA. We use the unitary matrix pencil method to estimate the frequencies of the waves, and use a least squares solver to estimate the amplitude and phase coefficients. We combine these to compute the DOA and evaluate the approach using simulations. We show that the method is very effective at estimating DOA for different numbers of incident waves, and different noise levels.

We describe an approach to elastically scale the performance of a data analytics operator that is part of a streaming application. Our techniques focus on dynamically adjusting the amount of computation an operator can carry out in response to changes in incoming workload and the availability of processing cycles. We show that our elastic approach is beneficial in light of the dynamic aspects of streaming workloads and stream processing environments. Addressing another recent trend, we show the importance of our approach as a means to providing computational elasticity in multicore processor-based environments such that operators can automatically find their best operating point. Finally, we present experiments driven by synthetic workloads, showing the space where the optimizing efforts are most beneficial and a radioastronomy imaging application, where we observe substantial improvements in its performance-critical section.

The least squares support vector machine (LS-SVM), like the SVM, is based on the margin-maximization performing structural risk and has excellent power of generalization. In this paper, we consider its use in semisupervised learning. We propose two algorithms to perform this task deduced from the transductive SVM idea. Algorithm 1 is based on combinatorial search guided by certain heuristics while Algorithm 2 iteratively builds the decision function by adding one unlabeled sample at the time. In term of complexity, Algorithm 1 is faster but Algorithm 2 yields a classifier with a better generalization capacity with only a few labeled data available. Our proposed algorithms are tested in several benchmarks and give encouraging results, confirming our approach.

This paper describes an application of the Minimum Classification Error (MCE) criterion to the problem of recognizing online unconstrained-style characters and words. We describe an HMM-based, character and word-level MCE training aimed at minimizing the character or word error rate while enabling flexibility in writing style through the use of multiple allographs per character. Experiments on a writer-independent character recognition task covering alpha-numerical characters and keyboard symbols show that the MCE criterion achieves more than 30 percent character error rate reduction compared to the baseline Maximum Likelihood-based system. Word recognition results, on vocabularies of 5k to 10k, show that MCE training achieves around 17 percent word error rate reduction when compared to the baseline Maximum Likelihood system.

This paper aims to improve the performance of an HMM-based offline Thai handwriting recognition system through discriminative training and the use of fine-tuned feature extraction methods. The discriminative training is implemented by maximizing the mutual information between the data and their classes. The feature extraction is based on our proposed block-based PCA and composite images, shown to be better at discriminating Thai confusable characters. We demonstrate significant improvements in recognition accuracies compared to the classifiers that are not discriminatively optimized.

In this paper we describe a new approach to dynamic signature verification using the discriminative training framework. The authentic and forgery samples are represented by two separate Gaussian Mixture models and discriminative training is used to achieve optimal separation between the two models. An enrollment sample clustering and screening procedure is described which improves the robustness of the system. We also introduce a method to estimate and apply subject norms representing the "typical" variation of the subject's signatures. The subject norm functions are parameterized, and the parameters are trained as an integral part of the discriminative training. The system was evaluated using 480 authentic signature samples and 260 skilled forgery samples from 44 accounts and achieved an equal error rate of 2.25%.

The paper describes an application of the prototype-based minimum error classifier (PBMEC) to the offline recognition of handwritten digits. The PBMEC uses a set of prototypes to represent each digit along with an L/sub /spl nu//-norm of distances as the decoding scheme. Optimization of the system is based on the minimum classification error (MCE) criterion. We introduce a new clustering criterion adapted to the PBMEC structure that minimizes an L/sub /spl nu//-norm-based distortion measure. The new clustering algorithm can generate a smaller number of prototypes than the standard k-means with no loss in accuracy. It is also shown that the PBMEC trained with MCE can achieve over 42% improvement from the baseline k-means process and requires only 28 Kb storage to match the performance of a 1.46 Mb sized k-NN classifier.

Discriminative feature extraction (DFE) has been proposed as a extension of MCE/GPD for the joint optimization of features and models. This study presents various configurations of this discriminative framework aimed at optimizing filter-bank parameters, using cepstrum and delta cepstrum as features, within an HMM-based system. Features and models are optimized either jointly or separately. Experimental results on the ISOLET database show that the joint optimization of features and models realizes the best performance: more than 13% absolute error rate reduction on the E-set task compared to an MLE-trained system using MFCCs and more than 1.85% absolute error rate reduction compared to an MCE-trained system using MFCCs.

This paper describes and evaluates the maximum mutual information criterion (MMI) for online unconstrained-style handwritten digit recognition based on hidden Markov models (HMMs). The study focuses on determining the best MMI optimization scheme and the HMM parameters that exhibit the most discriminative capabilities in the context of tied mixture density hidden Markov models (TDHMMs), where all HMM states share a pool of Gaussians. The experimental results show that the second-order optimization scheme is the most efficient and that although means and covariance matrix are shared by all models, they contribute the most to discrimination.

This paper proposes a model selection criterion for classification problems. The criterion focuses on selecting models that are discriminant instead of models based on the Occam's razor principle of parsimony between accurate modeling and complexity. The criterion, dubbed discriminative information criterion (DIC), is applied to the optimization of hidden Markov model topology aimed at the recognition of cursively-handwritten digits. The results show that DIC-generated models achieve 18% relative improvement in performance from a baseline system generated by the Bayesian information criterion (BIC).

This paper addresses the problem of estimating the optimal Hidden Markov Model (HMM) topology. The optimal topology is defined as the one that gives the smallest error-rate with the minimal number of parameters. The paper introduces a Bayesian model selection criterion that is suitable for Continuous Hidden Markov Models topology optimization. The criterion is derived from the Laplacian approximation of the posterior of a model structure, and shares the algorithmic simplicity of conventional Bayesian selection criteria, such as Schwarz's Bayesian Information Criterion (BIC). Unlike, BIC, which uses a multivariate Normal distribution assumption for the prior of all parameters of the model, the proposed HMM-oriented Bayesian Information Criterion (HBIC), models each parameter by a different distribution, one more appropriate for that parameter The results on an handwriting recognition task shows that the HBIC realizes a much smaller and efficient system than a system generated through the BIC.

This paper evaluates the application of minimum classification error (MCE) training to online-handwritten text recognition based on hidden Markov models. We describe an allograph-based, character level MCE training aimed at minimizing the character error rate while enabling flexibility in writing style. Experiments on a writer-independent discrete character recognition task, covering all alpha-numerical characters and keyboard symbols, show that MCE achieves more than 30% character error rate reduction compared to the baseline maximum likelihood-based system.

This paper addresses the problem of hidden Markov model (HMM) topology estimation in the context of on-line handwriting recognition. HMM have been widely used in applications related to speech and handwriting recognition with great success. One major drawback with these approaches, however, is that the techniques that they use for estimating the topology of the models (number of states, connectivity between the states and the number of Gaussians), are usually heuristically derived, without optimal certainty. This paper addresses this problem, by comparing a couple of commonly used heuristically derived methods to an approach that uses the Bayesian information criterion (BIC) for computing the optimal topology. Experimental results on discretely written letters show that using BIC gives comparable results to heuristic approaches with a model that has nearly 10% fewer parameters.

A pattern recognizer is usually a modular system which consists of a feature extractor module and a classifier module. Traditionally, these two modules have been designed separately, which may not result in an optimal recognition accuracy. To alleviate this fundamental problem, the authors have developed a design method, named discriminative feature extraction (DFE), that enables one to design the overall recognizer, i.e., both the feature extractor and the classifier, in a manner consistent with the objective of minimizing recognition errors. This paper investigates the application of this method to designing a speech recognizer that consists of a filter-hank feature extractor and a multi-prototype distance classifier. Carefully investigated experiments demonstrate that DFE achieves the design of a better recognizer and provides an innovative recognition-oriented analysis of the filter-bank, as an alternative to conventional analysis based on psychoacoustic expertise or heuristics.

This paper describes progress on a commercial application of the MECS recognition system to the task of recognizing Japanese family names spoken by customers into the answering machines of a large marketing/human resource company. The task is thus speaker-independent, open vocabulary, and is characterized by large variation in caller speaking styles, telephone types and acoustic environments. Our results show that context-independent hidden Markov models trained discriminatively with the minimum classification error criterion are a practical alternative to context-dependent models based on phonetic decision trees, yielding better performance with a much smaller number of parameters. On this difficult task we have obtained 59% correct family name recognition. A phoneme-based confidence measure enables us to obtain 85% correct name recognition for accepted utterances, at an overall utterance acceptance rate of 15%.

This paper investigates the realization of optimal filter bank-based cepstral parameters. The framework is the discriminative feature extraction method (DFE) which iteratively estimates the filter-bank parameters according to the errors that the system makes. Various parameters of the filter-bank, such as center frequency, bandwidth, and gain are optimized using a string-level optimization and a frame-level optimization scheme. Application to vowel and noisy telephone speech recognition tasks shows that the DFE method realizes a more robust classifier by appropriate feature extraction.

We propose a simple but powerful method for normalizing various sources of mismatch between training and testing conditions in speech recognizers, based on a training methodology called the generalized probabilistic descent method (GPD). In this new framework, a gradient based method is used to adapt the parameters of the feature extraction process in order to minimize the distortion between new speech data and existing classifier models, while most conventional normalization/adaptation methods attempt to adapt classification parameters. The GPD was proposed as a general discriminative training method for pattern recognizers such as neural networks. Up until now this has been used only for classifier design, sometimes in combination with the design of a non adaptive feature extractor. This paper, in contrast, studies the adaptive training benefits of GPD in the framework of normalizing the feature extractor to a new pattern environment. Experiments which use this technique to improve Japanese vowel classification were conducted and demonstrate the ability to reduce error rates by as much as 40%.

This paper investigates the realization of a filter bank model that achieves minimum classification error. A bank-of-filter feature extractor module is jointly optimized with the classifier 's parameters so as to minimize the errors occurring at the back-end classifier, in the framework of Minimum Classification Error /Generalized Probabilistic Descent Method (MCE/GPD). The method was first applied to readjusting various parameters of filter banks linearly spaced on the Mel-scale for the Japanese vowel recognition task. Analysis of the feature extraction process shows how those parts of the spectrum that are relevant to discrimination are captured. Then the method was applied to a multi-speaker word recognition system, which resulted in an word error rate reduction of more than 20 %. 1. INTRODUCTION It has been argued that the interaction between signal representation and classification strongly influences speech recognition performance [2], [6]. However, the underlying nature of this...

A filter bank model, which achieves minimum error, is investigated in this paper. A bank-of-filter feature extractor module is comprehensively optimized with the classifier's parameters for minimization of the errors occurring at the back-end classifier. The method has been applied to readjusting Mel-scale and Bark-scale based filter banks for the Japanese vowel recognition task, the framework being provided by the minimum classification error (MCE)/generalised probabilistic descent method (GPD). The results show suggestive phenomena underlying the accuracy of the proposed approach.>