Abstracto

Big Data Analytics using Meta Machine Learning

Bharati Suvalka ,Sarika kandelwal , Sidharth Singh Sisodia

As “Big Data” grows bigger at a rapid speed, Machine Learning (MACHINE LEARNING) techniques have come to play a vital role in automatic data processing and analytics across a wide spectrum of application domains. However, lack of well-defined values in choosing MACHINE LEARNING algorithms suitable for a given problem remains a major challenge. Today this choice depends mainly upon empirical rules such as the size of training data, number of dissimilar labels, need for interpretable decision boundaries, and real-time memory constraint. It is often also guided by realistic factors such as readily available code and comfort level of the programmers, and empirically unwavering parameters finely tuned by repeated experiments. We propose to lay the foundations of the next generation of domain agnostic MACHINE LEARNING techniques which will be able to sum up “a priori” knowledge of MACHINE LEARNING successes across domains, in analytics framework. Our goal is to alter the difficult alchemy involved in using MACHINE LEARNING techniques that take years to master into a simple skill that can be readily adapted by practitioners across fields. We wish to refer to this science as “Meta-Machine Learning” (MMACHINE LEARNING).