MCBoost: Multiple classifier boosting for perceptual co-clustering of images and visual features

We present a new co-clustering problem of images and visual features. The problem involves a set of non-object images in addition to a set of object images and features to be co-clustered. Co-clustering is performed in a way that maximises discrimination of object images from non-object images, thus emphasizing discriminative features. This provides a way of obtaining perceptual joint-clusters of object images and features. We tackle the problem by simultaneously boosting multiple strong classifiers which compete for images by their expertise. Each boosting classifier is an aggregation of weak-learners, i.e. simple visual features. The obtained classifiers are useful for object detection tasks which exhibit multimodalities, e.g. multi-category and multi-view object detection tasks. Experiments on a set of pedestrian images and a face data set demonstrate that the method yields intuitive image clusters with associated features and is much superior to conventional boosting classifiers in object detection tasks.

Enhancement of medical named entity recognition using graph-based features

Named Entity Recognition (NER) is a crucial step in text mining. This paper proposes a new graph-based technique for representing unstructured medical text. The new representation is used to extract discriminative features that are able to enhance the NER performance. To evaluate the usefulness of the proposed graph-based technique, the i2b2 medication challenge data set is used. Specifically, the 'treatment' named entities are extracted for evaluation using six different classifiers. The F-measure results of five classifiers are enhanced, with an average improvement of up to 26% in performance.

Physical activity recognition from accelerometer data using a multi-scale ensemble method

Accurate and detailed measurement of an individual's physical activity is a key requirement for helping researchers understand the relationship between physical activity and health. Accelerometers have become the method of choice for measuring physical activity due to their small size, low cost, convenience and their ability to provide objective information about physical activity. However, interpreting accelerometer data once it has been collected can be challenging. In this work, we applied machine learning algorithms to the task of physical activity recognition from triaxial accelerometer data. We employed a simple but effective approach of dividing the accelerometer data into short non-overlapping windows, converting each window into a feature vector, and treating each feature vector as an i.i.d training instance for a supervised learning algorithm. In addition, we improved on this simple approach with a multi-scale ensemble method that did not need to commit to a single window size and was able to leverage the fact that physical activities produced time series with repetitive patterns and discriminative features for physical activity occurred at different temporal scales.

Automated feature engineering for HTTP tunnel detection

Generating discriminative input features is a key requirement for achieving highly accurate classifiers. The process of generating features from raw data is known as feature engineering and it can take significant manual effort. In this paper we propose automated feature engineering to derive a suite of additional features from a given set of basic features with the aim of both improving classifier accuracy through discriminative features, and to assist data scientists through automation. Our implementation is specific to HTTP computer network traffic. To measure the effectiveness of our proposal, we compare the performance of a supervised machine learning classifier built with automated feature engineering versus one using human-guided features. The classifier addresses a problem in computer network security, namely the detection of HTTP tunnels. We use Bro to process network traffic into base features and then apply automated feature engineering to calculate a larger set of derived features. The derived features are calculated without favour to any base feature and include entropy, length and N-grams for all string features, and counts and averages over time for all numeric features. Feature selection is then used to find the most relevant subset of these features. Testing showed that both classifiers achieved a detection rate above 99.93% at a false positive rate below 0.01%. For our datasets, we conclude that automated feature engineering can provide the advantages of increasing classifier development speed and reducing development technical difficulties through the removal of manual feature engineering. These are achieved while also maintaining classification accuracy.

A confidence-based approach for improving keyword hypothesis scores

The task in keyword spotting (KWS) is to hypothesise times at which any of a set of key terms occurs in audio. An important aspect of such systems are the scores assigned to these hypotheses, the accuracy of which have a significant impact on performance. Estimating these scores may be formulated as a confidence estimation problem, where a measure of confidence is assigned to each key term hypothesis. In this work, a set of discriminative features is defined, and combined using a conditional random field (CRF) model for improved confidence estimation. An extension to this model to directly address the problem of score normalisation across key terms is also introduced. The implicit score normalisation which results from applying this approach to separate systems in a hybrid configuration yields further benefits. Results are presented which show notable improvements in KWS performance using the techniques presented in this work. © 2013 IEEE.

Automatic Analysis of Digital Retinal Images for Glaucoma Detection

In this paper we propose a novel automated glaucoma detection framework for mass-screening that operates on inexpensive retinal cameras. The proposed methodology is based on the assumption that discriminative features for glaucoma diagnosis can be extracted from the optical nerve head structures,
such as the cup-to-disc ratio or the neuro-retinal rim variation. After automatically segmenting the cup and optical disc, these features are feed into a machine learning classifier. Experiments were performed using two different datasets and from the obtained results the proposed technique provides
better performance than approaches based on appearance. A main advantage of our approach is that it only requires a few training samples to provide high accuracy over several different glaucoma stages.

Classification-based probabilistic modeling of texture transition for fast line search tracking and delineation

We introduce a classification-based approach to finding occluding texture boundaries. The classifier is composed of a set of weak learners, which operate on image intensity discriminative features that are defined on small patches and are fast to compute. A database that is designed to simulate digitized occluding contours of textured objects in natural images is used to train the weak learners. The trained classifier score is then used to obtain a probabilistic model for the presence of texture transitions, which can readily be used for line search texture boundary detection in the direction normal to an initial boundary estimate. This method is fast and therefore suitable for real-time and interactive applications. It works as a robust estimator, which requires a ribbon-like search region and can handle complex texture structures without requiring a large number of observations. We demonstrate results both in the context of interactive 2D delineation and of fast 3D tracking and compare its performance with other existing methods for line search boundary detection.

Prediction of age, sentiment, and connectivity from social media text

Social media corpora, including the textual output of blogs, forums, and messaging applications, provide fertile ground for linguistic analysis material diverse in topic and style, and at Web scale. We investigate manifest properties of textual messages, including latent topics, psycholinguistic features, and author mood, of a large corpus of blog posts, to analyze the impact of age, emotion, and social connectivity. These properties are found to be significantly different across the examined cohorts, which suggest discriminative features for a number of useful classification tasks. We build binary classifiers for old versus young bloggers, social versus solo bloggers, and happy versus sad posts with high performance. Analysis of discriminative features shows that age turns upon choice of topic, whereas sentiment orientation is evidenced by linguistic style. Good prediction is achieved for social connectivity using topic and linguistic features, leaving tagged mood a modest role in all classifications.

Development of a speaker recognition system using wavelets and artificial neural networks

This paper addresses the problem of speaker recognition from speech signals. The study focuses on the development of a speaker recognition system comprising two modules: a wavelet-based feature extractor, and a neural-network-based classifier. We have conducted a number of experiments to investigate the applicability of Discrete Wavelet Transform (D WT) in extracting discriminative features from the speech signals, and have examined various models from the Adaptive Resonance Theory (ART) family of neural networks in classijjing the extracted features. The results indicate that DWT could be a potential feature extraction tool for speaker recognition. In addition, the ART-based classijiers have yielded very promising recognition accuracy at more than 81%.

Human identification from ECG signals via sparse representation of local segments

This work proposes a novel framework to extract compact and discriminative features from Electrocardiogram (ECG) signals for human identification based on sparse representation of local segments. Specifically, local segments extracted from an ECG signal are projected to a small number of basic elements in a dictionary, which is learned from training data. A final representation is extracted by performing a max pooling procedure over all the sparse coefficient vectors in the ECG signal. Unlike most of existing methods for human identification from ECG signals which require segmentation of individual heartbeats or extraction of fiducial points, the proposed method does not need to segment individual heartbeats or detect any fiducial points. The method achieves an 99.48% accuracy on a 100 subjects dataset constructed from a publicly available database, which demonstrates that both local and global structural information are well captured to characterize the ECG signals.

Online social capital : mood, topical and psycholinguistic analysis

Social media provides rich sources of personal information and community interaction which can be linked to aspect of mental health. In this paper we investigate manifest properties of textual messages, including latent topics, psycholinguistic features, and authors' mood, of a large corpus of blog posts, to analyze the aspect of social capital in social media communities. Using data collected from Live Journal, we find that bloggers with lower social capital have fewer positive moods and more negative moods than those with higher social capital. It is also found that people with low social capital have more random mood swings over time than the people with high social capital. Significant differences are found between low and high social capital groups when characterized by a set of latent topics and psycholinguistic features derived from blogposts, suggesting discriminative features, proved to be useful for classification tasks. Good prediction is achieved when classifying among social capital groups using topic and linguistic features, with linguistic features are found to have greater predictive power than latent topics. The significance of our work lies in the importance of online social capital to potential construction of automatic healthcare monitoring systems. We further establish the link between mood and social capital in online communities, suggesting the foundation of new systems to monitor online mental well-being.

Classification of healthcare data using genetic fuzzy logic system and wavelets

Healthcare plays an important role in promoting the general health and well-being of people around the world. The difficulty in healthcare data classification arises from the uncertainty and the high-dimensional nature of the medical data collected. This paper proposes an integration of fuzzy standard additive model (SAM) with genetic algorithm (GA), called GSAM, to deal with uncertainty and computational challenges. GSAM learning process comprises three continual steps: rule initialization by unsupervised learning using the adaptive vector quantization clustering, evolutionary rule optimization by GA and parameter tuning by the gradient descent supervised learning. Wavelet transformation is employed to extract discriminative features for high-dimensional datasets. GSAM becomes highly capable when deployed with small number of wavelet features as its computational burden is remarkably reduced. The proposed method is evaluated using two frequently-used medical datasets: the Wisconsin breast cancer and Cleveland heart disease from the UCI Repository for machine learning. Experiments are organized with a five-fold cross validation and performance of classification techniques are measured by a number of important metrics: accuracy, F-measure, mutual information and area under the receiver operating characteristic curve. Results demonstrate the superiority of the GSAM compared to other machine learning methods including probabilistic neural network, support vector machine, fuzzy ARTMAP, and adaptive neuro-fuzzy inference system. The proposed approach is thus helpful as a decision support system for medical practitioners in the healthcare practice.

Medical data classification using interval type-2 fuzzy logic system and wavelets

This paper introduces an automated medical data classification method using wavelet transformation (WT) and interval type-2 fuzzy logic system (IT2FLS). Wavelet coefficients, which serve as inputs to the IT2FLS, are a compact form of original data but they exhibits highly discriminative features. The integration between WT and IT2FLS aims to cope with both high-dimensional data challenge and uncertainty. IT2FLS utilizes a hybrid learning process comprising unsupervised structure learning by the fuzzy c-means (FCM) clustering and supervised parameter tuning by genetic algorithm. This learning process is computationally expensive, especially when employed with high-dimensional data. The application of WT therefore reduces computational burden and enhances performance of IT2FLS. Experiments are implemented with two frequently used medical datasets from the UCI Repository for machine learning: the Wisconsin breast cancer and Cleveland heart disease. A number of important metrics are computed to measure the performance of the classification. They consist of accuracy, sensitivity, specificity and area under the receiver operating characteristic curve. Results demonstrate a significant dominance of the wavelet-IT2FLS approach compared to other machine learning methods including probabilistic neural network, support vector machine, fuzzy ARTMAP, and adaptive neuro-fuzzy inference system. The proposed approach is thus useful as a decision support system for clinicians and practitioners in the medical practice. copy; 2015 Elsevier B.V. All rights reserved.

Automatic visual dictionary generation through Optimum-Path Forest clustering

Image categorization by means of bag of visual words has received increasing attention by the image processing and vision communities in the last years. In these approaches, each image is represented by invariant points of interest which are mapped to a Hilbert Space representing a visual dictionary which aims at comprising the most discriminative features in a set of images. Notwithstanding, the main problem of such approaches is to find a compact and representative dictionary. Finding such representative dictionary automatically with no user intervention is an even more difficult task. In this paper, we propose a method to automatically find such dictionary by employing a recent developed graph-based clustering algorithm called Optimum-Path Forest, which does not make any assumption about the visual dictionary's size and is more efficient and effective than the state-of-the-art techniques used for dictionary generation. © 2012 IEEE.