Fuzzy Theoretic Data Models
This study introduces a fuzzy theoretic approach to deep learning such that fuzzy membership functions are used to study the propagation of uncertainties through the layers of a deep model. The most significance feature of the learning approach is that all of the unobserved variables and parameters associated to the deep model are averaged out where the averages are computed taking into account the uncertainties (on variables and parameters) being represented by the fuzzy membership functions optimally learned from the observed data. Several issues are addressed to design fast, competent, and robust modeling algorithms. A mathematical theory is provided to facilitate the application of fuzzy theoretic data models in prediction, filtering, and classification problems. The study is a theoretical contribution to the field of fuzzy machine learning nevertheless offering practical modeling algorithms.
- M. Kumar, S. Chatterjee, W. Zhang, and L. M. Kolbe, “Fuzzy Theoretic Data Models,” Information Sciences, under-preparation.