Prediction of Juvenile Delinquencies in Correlation with Education

Juvenile Crime is not a healthy trait for any developing country and certainly not for a widely cultured and populous country like India. India has been suffering from overpopulation and low literacy for many years now and it affects the growth of the nation. Education is the key to improve the state of the country and more juvenile crime, children receive lesser quality education and it keeps pulling down the nation. Not only that, it also poisons the next generation and is not healthy for the future of the nation. Even though education is provided for the convicted children, not everyone who is released continues to study which leads to nothing but more uneducated adults who are less productive in this competitive world. This project analyses the juvenile delinquency rate over the course of 30 Years and predicts the possible rate of juvenile crime in India in the forthcoming years, taking into account the various factors that would affect said juvenile crime rate, which include but not limited to: overpopulation, education especially between the ages 16 to 18, financial state of family, cultural differences, regional differences and past criminal history within the convicted juvenile's family. Since factoring in all the aforementioned features into account is near impossible, various machine learning (Linear Regression, Polynomial Regression regression, Back Propagation Network and Adam Optimization Algoritms) and deep learning models are created to primarily focus on the admission rates at different education levels in India. The results of said models provide us with insight on how important education is and how would the future generations shape up when the current trend continues, and also which of the selected learning models is the most accurate.