Improvement of Key Financial Performance Indicators in the Insurance Industry Using Machine Learning – A Quantitative Analysis.

Abstract

AI and Machine learning are playing a vital role in the financial domain in predicting future growth and risk and identifying key performance areas. We look at how machine learning and artificial intelligence (AI) directly or indirectly alter financial management in the banking and insurance industries. First, a non-technical review of the prior machine learning and AI methodologies beneficial to KPI management is provided. This paper will analyze and improve key financial performance indicators in insurance using machine learning (ML) algorithms. Before applying an ML algorithm, we must determine the attributes directly impacting the business and target attributes. The details must be manually mapped from string values to fit the model and its required datatypes for applying these specific features to an ML model. We propose hashing to convert string values to numeric values for data analysis within our model. After the string values are hashed, we can introduce our model. In our case, we have chosen to use a decision tree model. Decision Trees are beneficial for this use case as this algorithm generates rulesets that govern the target value output. These rulesets can then be applied to the financial dataset and infer the “best fit” value that might be wrong/missing. Finally, because of the model, we can use this most accurate data version to detect general ledger transactional data patterns.