No abstract

No abstract

The Minimal Learning Machine (MLM) has been recently proposed as a novel supervised learning method for regression problems aiming at reconstructing the mapping between input and output distance matrices. Estimation of the response is then achieved from the geometrical configuration of the output points. Thanks to its comprehensive formulation, the MLM is inherently capable of dealing with nonlinear problems and multidimensional output spaces. In this paper, we introduce an extension of the MLM to classification tasks, thus providing a unified framework for multiresponse regression and classification problems. On the basis of our experiments, the MLM achieves results that are comparable to many de facto standard methods for classification with the advantage of offering a computationally lighter alternative to such approaches. 1st BRICS Countries Congress on Computational Intelligence978-1-4799-3194-1/13 $31.00

Health insurance companies own very large databases built from the history of clinical exams and/or hospital procedures undergone by their beneficiaries. An important challenge faced by these companies is then to mine useful information from those database for the purpose of preventive care and financial costs reduction. Bearing this in mind, in this paper we propose a novel approach for building and labelling feature vectors for the beneficiaries of health insurance companies with the aim of building classifiers capable of predicting the risk level (high or low) of a given beneficiary to undergo serious cardiovascular events within a predefined horizon in the near future. The proposed approach was evaluated in the design of neural network classifiers using real-world health data from a Brazilian insurance company. The obtained results show that the proposed method is rather promising and can be used to aid the management of health insurance plans.

In this work, we discuss a recently proposed approach for supervised dimensionality reduction, the Supervised Distance Preserving Projection (SDPP) and, we investigate its applicability to monitoring material's properties from spectroscopic observations using Local Linear Regression (LLR). An experimental evaluation is conducted to show the performance of the SDPP and LLR and compare it with a number of state-of-theart approaches for unsupervised and supervised dimensionality reduction. For the task, the results obtained on a benchmark problem consisting of a set of NIR spectra of diesel fuels and six different chemico-physical properties of those fuels are discussed. Based on the experimental results, the SDPP leads to accurate and parsimonious projections that can be effectively used in the design of estimation models based on local linear regression.