Optimal feature weighting for the continuous HMM

In spotting for phrases in text images, speed and accuracy are important considerations. In a hidden Markov model (HMM) based spotter recognition time is dominated by the time required to compute the state conditional observation probabilities. These probabilities are a measure of how well the data match each state in the model. In this paper discrete and continuous hidden Markov models are compared based on speed and accuracy in spotting for phrases in text images. For the discrete HMM, vector quantization is used to associate each continuous feature vector with a discrete value. For the continuous HMMs, the observation distributions for the feature vectors are modeled by either a single Gaussian, or a mixture of two Gaussians. Comparisons were made on a subset of the UW English Document Image Database I. The best accuracy was observed when a mixture of two Gaussians was used in the continuous HMM. The discrete HMM provides for faster spotting particularly when long phrases are used.

Supervised pattern recognition for the prediction of contrast-enhancement appearance in brain tumors from multivariate magnetic resonance imaging and spectroscopy.

Simultaneous positive sequential vectors modeling and unsupervised feature selection via continuous hidden Markov models

In order to make Intelligent Transportation System (ITS) work effectively, a driver intention recognition method is proposed. In this research, three different recognition models were developed based on Continuous Hidden Markov Model (CHMM), and could distinguish left and right lane change intention from normal lane keeping intention. Subjects performed lane change maneuvers and lane keeping maneuvers with driving simulator which simulated highway scenes, parameters that highly correlated with lane change behavior were collected and analyzed. A series of testings and comparisons were done to obtain the optimal model structure and feature set. Results show that, taking the steering wheel angel, steering wheel angle velocity and lateral acceleration as the optimal observation signals, the accuracy can achieve up 95%, and it proved very effective in terms of early intention recognition.

This paper presents a parallel phoneme recognition algorithm based on the continuous Hidden Markov Model (HMM). The parallel phoneme recognition algorithm distributes 3-state HMMs of context dependent phonemes to the multiprocessors, computes output probabilities in parallel, and enhances the Viterbi beam search with a message passing mechanism. The algorithm is implemented in a multi-transputer system using distributed-memory MIMD multiprocessors. Experimental results show the feasibility of the parallel phoneme recognition algorithm in constructing a real-time parallel speech recognition system based on time-consuming continuous HMM.

We propose a modified discrete HMM that handles multimodalities. We assume that the feature space is partitioned into subspaces generated by different sources of information. To combine these heteregoneous modalities we propose a multi-stream discrete HMM that assigns a relevance weight to each subspace. The relevance weights are set local and depend on the symbols and the states. In particular, we associate a partial probability with each symbol in each subspace. The overall observation state probability is then computed as an aggregation of the partial probabilities and their objective relevance weights based on a linear combination. The Minimum Classification Error (MCE) objective based on the Gradient Probabilistic Descent (GPD) optimization algorithm is reformulated to derive the update equations for the relevance weights and the partial state probabilities. The proposed approach is validated using synthetic and real data sets. The results are shown to outperform the baseline discrete HMM that treats all streams equally important.

Multi metric functional connectivity analysis based on continuous hidden Markov model with application in early diagnosis of Alzheimer’s disease

Research conducted these last years in the field of spatial cognition report empirical findings that are difficult to account for with the traditional visual cognitive model of scene perception. One of the major contributions of these findings has been to invite rethinking scene perception, which would benefit from not being apprehended as centered mainly on the sensory modality considered. On the contrary, the Multisource model of scene perception developed by Intraub et al. offers an alternative theoretical framework considering visual perception as an act of spatial cognition, with spatial information at its core. According to this model, during the initial understanding of a view, the cognitive system would be elaborating a multisource representation, with spatial information constituting an egocentric framework that conveys to the observer a sense of the environment in which he/she is embedded. Scene representation would be organized around an amodal spatial structure combining different sources of information: a bottom-up and external source of information derived from different modalities (e.g., visual, haptic), as well as internal sources of high-level information (i.e., amodal, conceptual and contextual information). These different sources of information would work together to create a simulation of the likely environment, integrating the perceived view into a broader spatial context. Beyond rethinking scene perception, one of the advances of the model is to unify different fields of cognition apprehended until then in isolation. The current paper aims to present this model and some of the results it allows to account for.

A framework based on hidden Markov model with adaptive weighting for microcystin forecasting and early-warning

Image retrieval is the process of finding images in the database that are similar to the query image by measuring how close the feature values of the query image are to other images. Image retrieval is currently dominated by approaches that combine several different representations or features. The optimal weight of each feature is needed in combining the image features such as color features, texture features, and shape features. In this study, we use a multi-layer perceptron artificial neural network (MLP) method to obtain feature weights automatically and simultaneously look for optimal weights. The color moment is used to find nine color features, Gray Level Co-occurrence Matrix (GLCM) to find four texture features, and Hu Moment to find seven shape features totaling 20 neurons and all of these features become the input layer in our MLP network. Three neurons in output layers become the automatic weight of each feature. These weights are used to combine each feature's role in obtaining the relevant image. Euclidean Distance is used to measure similarity. The average precision values obtained using automatic feature weights are 93.94% for the synthetic dataset, 91.19% for the Coil-100 dataset, and 54.31% for the Wang dataset. These results have an average difference of 5.06% with the target so automatic feature weighting works well. This value is obtained at a hidden layer size of 11 and a learning rate of 0.1. In addition, the use of automatic feature weighting gives more accurate results compared to manual feature weighting.

Datasets that are gathered from multiple sources are called multi-view datasets. These types of datasets represent different sets of feature attributes to form different views. Advanced computing and information technology have made it possible to collect and store a massive amount of data. As more features are added to each view, the data becomes sparse, and analysis suffers from the curse of dimensionality. Exploring and integrating diverse information from different views have been the focus of many approaches to improve clustering performance. The relationships among distinct views should not only be explored and analyze but should also consider the emerging high-dimensionality of each view to further improve the clustering performance. Recently, Yang and Benjamin proposed a clustering algorithm called “Feature Weighted Reduction PCM (FW-R-PCM) that calculates feature weights to identify relevant features and consequently, eliminates features that are irrelevant to reduce feature dimension from the entire feature space. In this paper, we propose a feature-weighted multi-view possibilistic c-means (FW-R-MVPCM) clustering algorithm, which extends FW-R-PCM to consider clustering high-dimensional multi-view datasets and perform feature reduction simultaneously. Our proposed FW-R-MVPCM can effectively improve the clustering performance using a weighting scheme that identifies, and selects relevant features from each view, and eliminate the irrelevant features, thus reducing the dimension of features in each view. Experiments on real datasets are performed to analyze the theoretical behavior of FW-R-MVPCM and to show its usefulness and effectiveness. Furthermore, FW-R-MVPCM is compared with FW-R-PCM, and other multi-view clustering algorithms such as WMCFS, SWVF, W-MV-PCM-L2, and WV-Co-FCM.

An organism's phenotype can be influenced by maternal cues and directly perceived environmental cues, as well as by its genotype at polymorphic loci, which can be interpreted as a genetic cue. In fluctuating environments, natural selection favors organisms that efficiently integrate different sources of information about the likely success of phenotypic alternatives. In such situations, it can be beneficial to pass on maternal cues that offspring can respond to. A maternal cue could be based on environmental cues directly perceived by the mother but also partly on cues that were passed on by the grandmother. We have used a mathematical model to investigate how the passing of maternal cues and the integration of different sources of information evolve in response to qualitatively different kinds of temporal and spatial environmental fluctuations. The model shows that the passing of maternal cues and the transgenerational integration of sources of information readily evolve. Factors such as the degree of temporal autocorrelation, the predictive accuracy of different environmental cues, and the level of gene flow strongly influence the expression of adaptive maternal cues and the relative weights given to different sources of information. We outline the main features of the relation between the characteristics of environmental fluctuations and the adaptive systems of phenotype determination and compare these predictions with empirical studies on cue integration.

The development of epitope-based vaccines crucially relies on the ability to classify Human Leukocyte Antigen (HLA) molecules into sets that have similar peptide binding specificities, termed supertypes. In their seminal work, Sette and Sidney defined nine HLA class I supertypes and claimed that these provide an almost perfect coverage of the entire repertoire of HLA class I molecules. HLA alleles are highly polymorphic and polygenic and therefore experimentally classifying each of these molecules to supertypes is at present an impossible task. Recently, a number of computational methods have been proposed for this task. These methods are based on defining protein similarity measures, derived from analysis of binding peptides or from analysis of the proteins themselves. In this paper we define both peptide derived and protein derived similarity measures, which are based on learning distance functions. The peptide derived measure is defined using a peptide-peptide distance function, which is learned using information about known binding and non-binding peptides. The protein derived similarity measure is defined using a protein-protein distance function, which is learned using information about alleles previously classified to supertypes by Sette and Sidney (1999). We compare the classification obtained by these two complimentary methods to previously suggested classification methods. In general, our results are in excellent agreement with the classifications suggested by Sette and Sidney (1999) and with those reported by Buus et al. (2004). The main important advantage of our proposed distance-based approach is that it makes use of two different and important immunological sources of information-HLA alleles and peptides that are known to bind or not bind to these alleles. Since each of our distance measures is trained using a different source of information, their combination can provide a more confident classification of alleles to supertypes.

TFIDF algorithm was used for feature weighting in text classification. But the result of classification was not very well because of lack of class information in feature weighting. The known class information in the training set was used to improve the traditional TFIDF feature weight algorithm. Class distinction ability and class description ability were introduced, respectively expressed by inverse class frequency and term frequency in class, document frequency in class. A new feature weight algorithm based on class information, TF_IDT, was proposed. Naive Bayes classifier was used to test the algorithm. The precision, recall and F1 measure were significantly increased. Macro F1 measure raise by 6.46%. It was proved to be useful for improving text classification to use class information in feature weighting. In addition, the computational complexity of the proposed algorithm was lower and more suitable for use in fields of limited computing capability. Keywords-text classification; feature weight; inverse class frequency; term frequency in class; document frequency in class

Despite being a well‐known problem, feature weighting and feature selection are a major predicament for clustering. Most of the algorithms, which provide weighting or selection of features, require the number of clusters to be known in advance. On the other hand, the existing automatic clustering procedures that can determine the number of clusters are computationally expensive and often do not make a room for feature weighting or selection. In this paper, we propose a Gibbs sampling‐based algorithm for the Dirichlet process mixture model, which can determine the number of clusters and can also incorporate a near‐optimal feature weighting. We show that in the limiting case, the algorithm approaches a hard clustering procedure, which resembles minimization of an underlying clustering objective similar to weighted k‐means with an additional forfeit for the number of clusters and hence retains the simplicity of the Llyod's heuristics. To avoid the trivial solution of the resulting linear program, we include an additional entropic penalty on the feature weights. The proposed algorithm is tested on several synthetic and real‐life datasets. Through a detailed experimental analysis, we demonstrate the competitiveness of our proposal against the baseline as well as state‐of‐the‐art procedures for centre‐based high‐dimensional clustering.