Majority Vote and Decision Template Based Ensemble Classifiers Trained on Event Related Potentials for Early Diagnosis of Alzheimer's Disease

Abstract

With the rapid increase in the population of elderly individuals affected by Alzheimer's disease, the need for an accurate, inexpensive and non-intrusive diagnostic biomarker that can be made available to community healthcare providers presents itself as a major public health concern. The feasibility of EEG as such a biomarker has gained a renewed attention as several recent studies, including our previous efforts, reported promising results. In this paper we present our preliminary results on using wavelet coefficients of event related potentials along with an ensemble of classifiers combined with majority vote and decision templates.

Inteligência computacional aplicada à detecção de câncer de mama

“…O modelo de decisão adotado foi o majoritário, o qual corresponde a um dos mais simples e intuitivos modelos, em que basicamente, se escolhe a classe pela maioria dos votos do conjunto, que caso exista alguma razão para acreditar que certos classi cadores são melhores ou piores que outros utiliza-se uma votação ponderada (Stepenosky et al;2006), a Figura 5 demonstra seu funcionamento.…”

Random subspace method with class separability weighting

“…We select L of P * features randomly from the original c class with P ‐dimensional feature vector to train base‐learners, where P * < P and N is the number of instances. However, at this stage, one or more subsets may have low class separability, which damage the final decision in the majority voting stage (Stepenosky et al, ; Kotsiantis, ). Therefore, the RSM offers an elegant solution for large‐dimensional and noisy data classification with a possibility of subspaced poor features causing a drawback.…”