Browsing by Subject "Pattern recognition"
Now showing 1 - 4 of 4
Results Per Page
Sort Options
Item An EEG feature selection toolbox for EEGLAB in the matlab environment(2011-08) Kerr, Andy S; Baker, Mary C.; Pal, RanadipA complete system is proposed to generate features from raw EEG data and qausi-optimally reduce the feature set based on classification rates. Several default features are included for generating feature sets, and the feature set is qausi-optimally reduced using stepwise regression algorithms based on the classification of known classes. A plug-in known as the Feature Selection Toolbox was developed for the open source EEGLAB toolbox within the MATLAB environment to accomplish the goals of this thesis. Synchrony measures of the EEG are becoming more common as a means to establish network links and general comparisons of different areas of the brain. The four default features included in the Feature Selection Toolbox are average power, correlation coefficient, magnitude squared coherence, and phase synchrony index. An exhaustive search is impractical in finding an optimal subset of features as the computational time increases exponentially with the number of desired features in the optimal subset. Three stepwise regression feature selection algorithms are implemented to select the near optimal feature subset with a nearly linear increase in computational time as the maximum number of selected features increases. An example study comparing Alzheimer's Disease to Mild Cognitive Impairment and controls demonstrates the usefulness of the tools developed as part of this thesis. Also, the tradeoffs of different options in the Feature Selection Toolbox are assessed from the results of the algorithm in classifying the responses of individuals to two different cognitive tasks, one involving visual stimulus and counting, the other involving visual stimulus and spatial reasoning.Item Aptamers as cross-reactive receptors : using binding patterns to discriminate biomolecules(2013-05) Stewart, Sara, 1980-; Anslyn, Eric V., 1960-; Ellington, Andrew D.Exploration into the use of aptamers as cross-reactive receptors was the focus of this work. Cross-reactivity is of interest for developing assays to identify complex targets and solutions. By exploiting the simple chemistries of aptamers, we hope to introduce a new class of receptors to the science of molecular discrimination. This manuscript first addresses the use designed aptamers for the identification of variants of HIV-1 reverse transcriptase. In this research aptamers were immobilized on a platform and were used to discriminate four variants of HIV-1 reverse transcriptase. It was found that not only could the array discriminate HIV-1 reverse transcriptase variants for which aptamers were designed, it would also discriminate variants for which no aptamers exist. A panel of aptamers was used to discriminate four separate cell lines, which were chosen as examples of complex targets. This aptamer panel was used to further explore the use of aptamers as cross-reactive sensors. Forty-six aptamers were selected from the literature that were designed to be specific to cells or molecules expected to be in the surface of cells. This panel showed differential binding patterns to each of the cell types, displaying cross-reactive behavior. During the course of this research, we also developed a novel ratiometric method of using aptamer count derived from next-generation sequencing as a method for discrimination. This is in lieu of the more commonly used fluorescent signals. Finally the use of multiple signals for pattern recognition routines was further explored by running various models using artificial data. Various situations were applied to replicate different possible situation which might arise when working with macromolecular interactions. The purpose of this was to advance the communities understanding and ability to interpret results from the pattern recognition methods of PCA and LDA.Item Automated Pattern Recognition for Intonation (PRInt) : an essay on intonational phonology and categorization(2012-12) Bacuez, Nicholas; Montreuil, Jean-Pierre; Blyth, Carl; Bullock, Barbara; Erk, Katrin; Smiljanic, RajkaThis dissertation provides experimental evidence for the validity of an intonational phonology. The widely used Autosegmental-Metrical theory con- tends that the phonological structure of intonation can be expressed with two tonal targets (L/H tones and derivatives) and retrieved from its phonetic im- plementations. However, it has not been specifically demonstrated so far in a systematic way. This dissertation argues that this view on intonational phonol- ogy considers the phonetic forms of intonation as instances of phonologically structured intonational units forming functionally discrete categories (tones and derivatives). The model of Pattern Recognition for Intonation (PRInt) applies the concepts of categorization (vagueness, prototype, degrees of typicality) to in- tonation in order to abstract the phonological structure of intonational cate- gories from the ranking, by degree of typicality, of their variations in phonetic implementation. First, instances belonging to an intonation category are collected. Sec- ond, a pattern recognition module, relying on the 4-layer structure protocol, extracts a feature vector from the phonetic data of each instance: a sequence of structurally organized tones (L/H tones and derivatives). Third, a fuzzy classifier, using two functions (frequency and similar- ity), organizes the data from the feature vectors of all instances by degree of typicality (grade of membership of values in multisets) and generates the phonological structure of the intonation category, the prototypical pattern, ex- tracted from all instances, and that subsumes them all. It also re-creates the phonetic implementations of the phonological structure but with their features ranked by degree of typicality. This allows the model to distinguish phono- logically distinct structures from phonetic variations of the same phonological structure. The model successfully extracted the phonological intonation structure associated to three modalities of closed questions in French: neutral, doubt- ful, and surprised. It found that neutral and doubtful closed questions are phonologically distinct while surprise is a phonetic allocontour of the neutral modality, in line with prior characterizations of these patterns. It demon- strated that a bi-tonal phonological structure of intonation can be retrieved from phonetic variations. A versatile modeling tool, PRInt will be developed to use its acquired knowledge to evaluate the categorical status of novel instances and to extract multiple phonological units from mixed corpora.Item Sensor-based machine olfaction with neuromorphic models of the olfactory system(Texas A&M University, 2007-04-25) Raman, BaranidharanElectronic noses combine an array of cross-selective gas sensors with a pattern recognition engine to identify odors. Pattern recognition of multivariate gas sensor response is usually performed using existing statistical and chemometric techniques. An alternative solution involves developing novel algorithms inspired by information processing in the biological olfactory system. The objective of this dissertation is to develop a neuromorphic architecture for pattern recognition for a chemosensor array inspired by key signal processing mechanisms in the olfactory system. Our approach can be summarized as follows. First, a high-dimensional odor signal is generated from a chemical sensor array. Three approaches have been proposed to generate this combinatorial and high dimensional odor signal: temperature-modulation of a metal-oxide chemoresistor, a large population of optical microbead sensors, and infrared spectroscopy. The resulting high-dimensional odor signals are subject to dimensionality reduction using a self-organizing model of chemotopic convergence. This convergence transforms the initial combinatorial high-dimensional code into an organized spatial pattern (i.e., an odor image), which decouples odor identity from intensity. Two lateral inhibitory circuits subsequently process the highly overlapping odor images obtained after convergence. The first shunting lateral inhibition circuits perform gain control enabling identification of the odorant across a wide range of concentration. This shunting lateral inhibition is followed by an additive lateral inhibition circuit with center-surround connections. These circuits improve contrast between odor images leading to more sparse and orthogonal patterns than the one available at the input. The sharpened odor image is stored in a neurodynamic model of a cortex. Finally, anti-Hebbian/ Hebbian inhibitory feedback from the cortical circuits to the contrast enhancement circuits performs mixture segmentation and weaker odor/background suppression, respectively. We validate the models using experimental datasets and show our results are consistent with recent neurobiological findings.