BLIGHT TICKET PREDICTION

This project is a real life situation happended in Detroit a city in Michigan.Every year the Detroit city issues fines in millions of dollars to people who allow the property to remain in deteroitaed condition and many of these remain unpaid.Blight ticket is the ticket issued as fine to residents who violate maintanance of properties.Here the target variable is called as "Compliance".My target is to predict the complaince i.e whether the blight ticket will be paid on time or not. So in order to do that,two data sets have been loaded and analyzing the features we can see several features that are not necessary for our training and includes some nan and unique values which has to be removed.The data cleaning has been done by analyzing the importance of features that are related to the target variable Grid searchCV method and random forest regressor has been used to predict the compliance and see which gives better score and efficiency.