This program implements linear regression and ridge regression (without the use of machine learning libraries) to analyze local crime statistics for 1,994 US communities. The response y is the crime rate. The data is analyzed using the closed form solution and gradient desent for both regression models. K-fold cross validation is used to find an optimum hyperparameter for ridge regression.
rachgoldstein / crime-rate-prediction Goto Github PK
View Code? Open in Web Editor NEWLogistic regression model to predict crime statistics