@inbook {418, title = {A statistical-genetic algorithm to select the most significant features in mammograms}, booktitle = {Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)}, volume = {4673 LNCS}, year = {2007}, pages = {189-196}, abstract = {

An automatic classification system into either malignant or benign microcalcification from mammograms is a helpful tool in breast cancer diagnosis. From a set of extracted features, a classifying method using neural networks can provide a probability estimation that can help the radiologist in his diagnosis. With this objective in mind, this paper proposes a feature selection algorithm from a massive number of features based on a statistical distance method in conjunction with a genetic algorithm (GA). The use of a statistical distance as optimality criterion was improved with genetic algorithms for selecting an appropriate subset of features, thus making this algorithm capable of performing feature selection from a massive set of initial features. Additionally, it provides a criterion to select an appropriate number of features to be employed. Experimental work was performed using Generalized Softmax Perceptrons (GSP), trained with a Strict Sense Bayesian cost function for direct probability estimation, as microcalcification classifiers. A Posterior Probability Model Selection (PPMS) algorithm was employed to determine the network complexity. Results showed that this algorithm converges into a subset of features which has a good classification rate and Area Under Curve (AUC) of the Receiver Operating Curve (ROC). {\^A}{\textcopyright} Springer-Verlag Berlin Heidelberg 2007.

}, keywords = {Breast cancer, Diagnosis, Feature extraction, Genetic algorithms, Mammography, Microcalcification classification, Network complexity, Neural network classifiers, Neural networks, Tumors}, isbn = {9783540742715}, issn = {03029743}, url = {http://www.scopus.com/inward/record.url?eid=2-s2.0-38149142403\&partnerID=40\&md5=ef139db3a0e5d603c4f721316abdcf2c}, author = {G V Sanchez-Ferrero and J I Arribas} } @article {420, title = {A model selection algorithm for a posteriori probability estimation with neural networks}, journal = {IEEE Transactions on Neural Networks}, volume = {16}, year = {2005}, pages = {799-809}, abstract = {

This paper proposes a novel algorithm to jointly determine the structure and the parameters of a posteriori probability model based on neural networks (NNs). It makes use of well-known ideas of pruning, splitting, and merging neural components and takes advantage of the probabilistic interpretation of these components. The algorithm, so called a posteriori probability model selection (PPMS), is applied to an NN architecture called the generalized softmax perceptron (GSP) whose outputs can be understood as probabilities although results shown can be extended to more general network architectures. Learning rules are derived from the application of the expectation-maximization algorithm to the GSP-PPMS structure. Simulation results show the advantages of the proposed algorithm with respect to other schemes. {\^A}{\textcopyright} 2005 IEEE.

}, keywords = {Algorithms, Automated, Biological, Breast Neoplasms, Computer simulation, Computer-Assisted, Computing Methodologies, Decision Support Techniques, Diagnosis, Estimation, Expectation-maximization, Generalized Softmax Perceptron (GSP), Humans, Mathematical models, Model selection, Models, Neural Networks (Computer), Neural networks, Numerical Analysis, Objective function, Pattern recognition, Posterior probability, Probability, Statistical, Stochastic Processes, algorithm, article, artificial neural network, automated pattern recognition, biological model, breast tumor, classification, cluster analysis, computer analysis, computer assisted diagnosis, decision support system, evaluation, human, mathematical computing, methodology, statistical model, statistics}, issn = {10459227}, doi = {10.1109/TNN.2005.849826}, url = {http://www.scopus.com/inward/record.url?eid=2-s2.0-23044459586\&partnerID=40\&md5=f00e7d86a625cfc466373a2a938276d0}, author = {J I Arribas and Jes{\'u}s Cid-Sueiro} }