Machine Learning Models for Predictive Analytics in Personal Finance

Abstract

AbstractMachine learning is an application of artificial intelligence where statistical data is processed by various algorithms that are generally automated in order to produce insights and inferences. It finds many applications in the field of personal finance for portfolio analysis, recommendation engines and even financial forecasting tools. Personal finance management is absolutely crucial to attain financial freedom as well as security. Long-term fiscal planning can provide a contingency against uncertainty as well as promote financial stability. The main objective of this paper is to propose RNN-based predictive model for personal finance. This paper provides a comprehensive analysis technique that can be utilized to manage the key financial parameters for an individual using machine learning models. In the proposed work, we have implemented three models: a linear regression model for expenditure prediction, RNN-based for stock prediction and logistic regression for retirement prediction. These models are trained and tested on the basis of both the individual user’s data and external data pertaining to the economy and the financial markets. In the proposed work, experimental results show an accuracy score of 83.55% for linear regression, 86.7% for RNN and 84.53% for logistic regression, each of which is used for a different phase of the proposed system.KeywordsMachine learningPersonal financeLinear regressionRecurrent neural networksLogistic regressionRetirement prediction