In this thesis, we study the least trimmed squares artificial neural networks (LTS-ANNs), which are generalization of the least trimmed squares (LTS) estimators frequently used in robust linear parametric regression problems to nonparametric artificial neural networks (ANNs) used for nonlinear regression problems. Two training algorithms are proposed in this thesis. The first algorithm is the incremental gradient descent algorithm. In order to speed up the convergence, the second training algorithm is proposed based on recursive least squares (RLS). Three illustrative examples are provided to test the performances of robustness against outliers for the classical ANNs and the LTS-ANNs. Simulation results show that upon proper selection of the trimming constant of the learning machines, LTS-ANNs are quite robust against outliers compared with the classical ANNs.

Trace amines (TAs), tryptamine, tyramine, octopamine, and beta-phenylethylamine, named for their low endogenous concentrations in mammals, are related to the classical monoamine transmitters, but have been understudied and thought of as false transmitters. They share structural, physiological, pharmacological, and metabolic similarities with the monoamines, including synthesis by the aromatic-L-amino acid decarboxylase (AADC) enzyme. In 2001, a new class of receptors preferentially activated by the TAs, termed trace amine-associated receptors (TAARs), was discovered establishing a mechanism for TA actions independent of classic monoaminergic mechanisms. While the TAs and some of their receptors are present in the mammalian central nervous system (CNS), their physiologic role remains uncertain. I hypothesized that the TAs are found intrinsically in the spinal cord, and that they are able to modulate spinal neural networks. Using immunohistochemistry, numerous spinal neurons were identified that express AADC, TAs, and TAARs. Similar results were seen for AADC and TAAR1 with in situ hybridization. The most consistent observation was for labeling D cells associated with the central canal and in motoneurons. Overall, these results provided evidence for the presence of an anatomical substrate onto which the TAs could have intrinsic biological actions in the spinal cord. Using exogenous application of the TAs in the isolated spinal cord in vitro, and in vivo in the mid-thoracic chronically spinalized, I showed that the TAs could induce rhythmic locomotor-like activity. TA injection-induced hindlimb motor rhythms observed in chronic spinalized animals, supports TA spinal actions independent of the descending monoaminergic systems. In the presence of NMDA, TA applications recruited a variety of rhythmic motor patterns in the isolated spinal cord. This ranged from locomotor activity indistinguishable from 5-HT/NMDA induced locomotion to complex patterns including, an episodic form of locomotion where there were locomotor bouts with intervening quiescent periods. TA actions of pattern generating circuits had slower kinetics of activation than 5-HT and NA, were attenuated in the presence of monoamine transport inhibitors, and had increased intracellular labeling when incubated in a nominally Na+-free solution. Together these results suggest that the TAs require transport into neurons to exert their actions, and that transport occurs by Na+-dependent monoamine transporters as well as Na+-independent transporters. Finally, I used the in vitro isolated spinal cord with attached hindlimbs to record electromyographic (EMG) activity from various hindlimb muscles to compare the relationship between the TAs and serotonin (5-HT) evoked motor coordination and to examine the ability of the TAs to modulate ongoing 5-HT and NMDA locomotor-like activity. The TAs produced both the continuous and episodic patterns on muscles as observed in ventral root recordings, but EMG recordings provided more detailed insight into specific muscle actions. The TAs also generally increased both frequency and amplitude of ongoing 5-HT locomotor frequency, with tyramine and octopamine also particularly able to alter 5-HT motor coordination patterns.

<p>Today one of the greatest aims within the area of bioinformatics is to gain a complete understanding of the functionality of genes and the systems behind gene regulation. Regulatory relationships among genes seem to be of a complex nature since transcriptional control is the result of complex networks interpreting a variety of inputs. It is therefore essential to develop analytical tools detecting complex genetic relationships.</p><p>This project examines the possibility of the data mining technique artificial neural network (ANN) detecting regulatory relationships between genes. As an initial step for finding regulatory relationships with the help of ANN the goal of this project is to train an ANN to predict the expression of an individual gene. The genes predicted are the nuclear receptor PPAR-g and the insulin receptor. Predictions of the two target genes respectively were made using different datasets of gene expression data as input for the ANN. The results of the predictions of PPAR-g indicate that it is not possible to predict the expression of PPAR-g under the circumstances for this experiment. The results of the predictions of the insulin receptor indicate that it is not possible to discard using ANN for predicting the gene expression of an individual gene.</p>

<p>The aim of this thesis is to describe how a statisticallybased neural network technology, here named BCPNN (BayesianConfidence Propagation Neural Network), which may be identifiedby rewriting Bayes' rule, can be used within a fewapplications, data mining and classification with credibilityintervals as well as unsupervised pattern recognition.</p><p>BCPNN is a neural network model somewhat reminding aboutBayesian decision trees which are often used within artificialintelligence systems. It has previously been success- fullyapplied to classification tasks such as fault diagnosis,supervised pattern recognition, hiearchical clustering and alsoused as a model for cortical memory. The learning paradigm usedin BCPNN is rather different from many other neural networkarchitectures. The learning in, e.g. the popularbackpropagation (BP) network, is a gradient method on an errorsurface, but learning in BCPNN is based upon calculations ofmarginal and joint prob- abilities between attributes. This isa quite time efficient process compared to, for instance,gradient learning. The interpretation of the weight values inBCPNN is also easy compared to many other networkarchitechtures. The values of these weights and theiruncertainty is also what we are focusing on in our data miningapplication. The most important results and findings in thisthesis can be summarised in the following points:</p><p>    We demonstrate how BCPNN (Bayesian Confidence PropagationNeural Network) can be extended to model the uncertainties incollected statistics to produce outcomes as distributionsfrom two different aspects: uncertainties induced by sparsesampling, which is useful for data mining; uncertainties dueto input data distributions, which is useful for processmodelling.</p><p>    We indicate how classification with BCPNN gives highercertainty than an optimal Bayes classifier and betterprecision than a naïve Bayes classifier for limited datasets.</p><p>    We show how these techniques have been turned into auseful tool for real world applications within the drugsafety area in particular.</p><p>    We present a simple but working method for doingautomatic temporal segmentation of data sequences as well asindicate some aspects of temporal tasks for which a Bayesianneural network may be useful.</p><p>    We present a method, based on recurrent BCPNN, whichperforms a similar task as an unsupervised clustering method,on a large database with noisy incomplete data, but muchquicker, with an efficiency in finding patterns comparablewith a well known (Autoclass) Bayesian clustering method,when we compare their performane on artificial data sets.Apart from BCPNN being able to deal with really large datasets, because it is a global method working on collectivestatistics, we also get good indications that the outcomefrom BCPNN seems to have higher clinical relevance thanAutoclass in our application on the WHO database of adversedrug reactions and therefore is a relevant data mining toolto use on the WHO database.</p><p>Artificial neural network, Bayesian neural network, datamining, adverse drug reaction signalling, classification,learning.</p>

Thesis (Ph. D.)--Texas A&M University, 2004. / Vita. Includes bibliographical references (p. 171-177).

Thesis (Ph. D.)--Lehigh University, 2000. / Includes vita. Includes bibliographical references (leaves 138-147).

Thesis (Ph.D.)--University of Canterbury, 2009. / Typescript (photocopy). "29th April 2009." Includes bibliographical references (p. 256-277). Also available via the World Wide Web.

Thesis (M.S.)--Rochester Institute of Technology, 2009. / Typescript. Includes bibliographical references (p. 120-126).

Thesis (M.S.) -- New Jersey Institute of Technology, Dept. of Computer and Information Science, 2000. / Includes bibliographical references. Also available via the World Wide Web.

Thesis (M.S.)--West Virginia University, 2002. / Title from document title page. Document formatted into pages; contains xii, 94 p. : ill. (some col.), maps. Includes abstract. Includes bibliographical references (p. 51-53).