Mathematics and Computer Science, Volume 1

13Data Representation and Performance in a Prediction Model

Apurbalal Senapati¹, Soumen Maji²* and Arunendu Mondal³

¹Department of Computer Science & Engineering, Central Institute of Technology Kokrajhar, Kokrajhar, Assam, India

²Department of Civil Engineering, Central Institute of Technology Kokrajhar, Kokrajhar, Assam, India

³Department of Chemistry, Central Institute of Technology Kokrajhar, Kokrajhar, Assam, India

Abstract

Prediction is a fundamental task in science and engineering disciplines, especially in numerical methods, data science, machine learning, weather forecasting, etc. There are several algorithms for predictions that are developed, also known as predictive modeling or predictive analytics. It is a mathematical model that attempts to predict future events or outcomes by analyzing the patterns of past observation. In a nutshell, it can be stated that, based on known past behavior, tried to answer what is most likely to happen in the future? The key element of future prediction is the past observation or past behavior is termed as past data or data. The data must be in a specific format such that it is fitted with a specific predictive model. But in real life, data are available in different forms and formats. Sometimes data may not be fitted into the various mathematical models directly or sometimes gives poor performance. This manuscript gives a comparative study of a predictive model with a different representation of the same data from which it can be ...