An approach for remove missing values in numerical and categorical values using two way table marginal joint probability

Abstract

Data analytics is a wide area which helps to extract the essential information from a huge volume of data. Data gathered from different sources are not in the same format, hence it is very difficult for pre-processing such data. Each and every data set has different categories of data types such as numerical or categorical data types. In this proposed work, we will discuss about identifying the missing values from the dataset using statistical techniques such as two-way table joint probability, two-way table marginal probabilities and two-way table conditional probability. The work also focuses on how to extract the essential features from the data set. Different visualization methods are used to easily understand the data set and feature prediction. © 2020 SERSC.