Incremental learning for large scale classification systems

Abstract

One of the main characteristics of our time is the growth of the data volumes. We collect data literally from everywhere; smart phones, smart devices, social media and the health care system, which defines a small portion of the sources of the big data. The big data growth poses two main difficulties, storing and processing them. For the former, there are certain new technologies that enable us to store large amounts of data in a fast and reliable way. For the latter, new application frameworks have been developed. In this paper, we perform classification analysis using Apache Spark in one real dataset. The classification algorithms that we have used are multiclass, and we are going to examine the effect of the dataset size and input features on the classification results.