Analysis of Suicide Victim Data for the Prediction of Number of Suicides in India

摘要

A study is presented aimed at discovering the prime factors that affect the number of suicides in certain regions of India in the year 2011 and subsequently using them to predict the number of suicides in the future. This prediction of suicides can help in making governing decisions in the affected regions such as promoting education and reworking existing facilities. The features in the study refer primarily to the part of the population which are affected most by suicides. Measures can be taken to reduce the number of suicides in the regions by setting up suicide preventive centers and targeting the affected population with special suicide preventive measures such as psychiatric counselling. The Indian government maintains a database of the reported suicide cases in each region of India; this is made public for the purpose of data analytics. Along with number of suicides for each of the considered states, the demo-graphical features of that state were used as features for creation of the model. The three categories of the dataset are census data, marital status data and educational status data of the states. Pearson correlation was computed to determine the strength of the features on the number of suicides and then linear regression was used to develop a model for the prediction of number of suicides. The results obtained were remarkable, 9 features were found to have a significant linear relationship with number of suicides. The prediction model developed using these 9 features gave us a linear fit close to 99.8% and prediction accuracy of 99.1%. Hence from this study a more meaningful insight on which section of the Indian population is most affected by suicides was obtained, this is an invaluable statistic for focusing preventive measures and making major governing decisions for different regions of the country. The number of suicides for anytime in the future can be predicted by the linear model developed here, which can serve as a measure for evaluating the effectiveness of different policy decisions and also that of various suicide preventive measures that are undertaken by NGOs or the government.