Computational Eco-Systems for Handwritten Digits Recognition

Abstract

Inspired by the importance of diversity in biological system, we built an heterogeneous system that could achieve this goal. Our architecture could be summarized in two basic steps. First, we generate a diverse set of classification hypothesis using both Convolutional Neural Networks, currently the state-of-the-art technique for this task, among with other traditional and innovative machine learning techniques. Then, we optimally combine them through Meta-Nets, a family of recently developed and performing ensemble methods.

Withdrawn from Journal for flaws in methodology

References in corpus (4)

• Deep Learning in Neural Networks: An Overview
• Estimating Continuous Distributions in Bayesian Classifiers
• Generalizing Pooling Functions in Convolutional Neural Networks: Mixed, Gated, and Tree
• Batch-normalized Maxout Network in Network

Cited by in corpus (1)

• MNIST-NET10: A heterogeneous deep networks fusion based on the degree of certainty to reach 0.1 error rate. Ensembles overview and proposal