An unsupervised feature selection method is proposed for analysis of data sets of high dimensionality. The Least Square Error (LSE) of approximating the complete data set via a reduced feature subset is proposed as the quality measure for feature selection. Guided by this LSE criterion, a feature selection algorithm is developed to find the feature subset with the lowest LSE. The algorithm (named KLS-FS) is granted the capability of non-linear feature selection by using the kernel representation. An incremental LSE computation is designed to accelerate the selection process and, therefore, enhances the scalability of KLS-FS to high-dimensional datasets. The superiority of the proposed feature selection algorithm, in terms of keeping principal data structures, learning performances in classification and clustering applications, and robustness, is demonstrated using various real-life datasets of different sizes and dimensions.
Liu R, Rallo R, Cohen Y. (2011) Unsupervised Feature Selection using Incremental Least Squares. International Journal of Information and Decision Making, 10(6):967-987
- PUBMED ID: N/A
Prediction of Modes of Toxic Action and toxicity of phenols with feature selection algorithms coupled with fuzzy ARTMAP
Descriptors suitable for the discrimination of four modes of toxic action (MOA) of 221 phenols with respect to the ciliate Tetrahymena pyriformis were selected by using SOM-dissimilarity measures and other well-known feature selection techniques. The performance of these methods and of ensembles of them to predict MOA classes and the toxicity log[1/IGC(50)] (mmol/ L) for each action mechanism when coupled with a fuzzy-ARTMAP neural network were assessed.
Rallo, R.,Espinosa, G., and F. Giralt. Prediction of Modes of Toxic Action and toxicity of phenols with feature selection algorithms coupled with fuzzy ARTMAP Proceedings of the Joint Meeting on Medicinal Chemistry, Vienna, Eds. P. Ettmayer and G. Ecker, Medimont SRL, F620C0176, 27-33. (ISBN 88-7587-163-9)
A predictive Fuzzy ARTMAP neural system and two hybrid networks, each combining a dynamic unsupervised classifier with a different kind of supervised mechanism, were applied to develop virtual sensor systems capable of inferring the properties of manufactured products from real process variables. A new method to construct dynamically the unsupervised layer was developed. A sensitivity analysis was carried out by means of self-organizing maps to select the most relevant process features and to reduce the number of input variables into the model. The prediction of the melt index (MI) or quality of six different LDPE grades produced in a tubular reactor was taken as a case study. The MI inferred from the most relevant process variables measured at the beginning of the process cycle deviated 5% from on-line MI values for single grade neural sensors and 7% for composite neural models valid for all grades simultaneously.
Rallo R, Ferré-Giné J, Arenas A, Giralt F (2002) Neural Virtual Sensor for the Inferential Prediction of Product Quality from Process Variables. Computers and Chemical Engineering (26) 12, 1735-54