Insurance Data Analysis with COGNITO: An Auto Analysing and Storytelling Python Library

摘要

Data pre-processing has taken an enhanced role with the advent of Machine learning. It is a vital element that forms the encore of the data science and business analytics process. Data pre-processing involves generating descriptive statistical summary, data cleaning, and data manipulation based on inputs gained after the initial analysis. Of late, it has been observed that data science practitioners spend 45% to 50% of their time cleaning and processing the data. Much time can be saved if the data transformation process can be automated. The COGNITO framework helps in performing the automated feature engineering and data storytelling of the dataset based on end-user discretion. The present work discusses the process and results obtained when automated feature engineering was performed on an insurance dataset using COGNITO.