Statistical Learning Theoretical Foundations Overview for Big Data Predictive Analytic

This paper presents a learning machine overview for Big Data Predictive Analytic. Produced data, in this decade, become bigger and bigger than ever. They have to be analysed and processed in order to extract relevant knowledge to make predictive analytic. Learning machines comes at this stage to estimate predictors based on observed historical data. Learning algorithms performance and data quantity evolution must be parallel to keep tolerable performance. This parallelism is one of main challenges of Big Data field. For that reason, this work introduces the basic theoretical foundations of learning machines to push researchers to design new algorithms taking the data amount and performance aspect in consideration