Deep Learning With Conceptual View in Meta Data for Content Categorization
Asokan Ramasamy, P. Preethi
Abstract
Data gathered from various devices have to be observed by human operators manually for extended durations which is not viable and may lead to imprecise results. Data are analyzed only when any unwanted event occurs. Machine-learning technology powers many aspects of modern society, from web searches to content filtering on social networks to recommendations on e-commerce websites, and it is increasingly present in consumer products. Machine-learning systems are used to identify objects in different forms of data. For decades, constructing a pattern-recognition, machine-learning system required careful engineering and domain expertise to design a feature extractor that transformed the raw data into a suitable internal representation, which the learning subsystem could detect patterns in the input by making use of and integrating ideas such as backpropagation, regularization, the softmax function, etc. This chapter will cover the importance of representations and metadata appendage and feature vector construction for the training deep models optimization.