The following code was developed by Dario Molinari, Daniele De Cillis, Lorenzo Vitali for the exam of Data Mining in the Master of Artificial Intelligence and Robotics of Sapienza University.
The aim of this project is to create a classifier able to predict a rating class of a future movie, submitting the actors, the director, the writer.
Moreover to understand the level of classification we compared our method with a another regression algorithm, also there are some codes regarding data analysis as "best actor by genre" , "most popular movie by genre".
Another statistics is made to implement a trend prediction, using the arima model and predicting the parameters reducing the error.
MAIN FILE DESCRIPTIONs :
LogisticRegression.py -> In this file is implemented the softmax regression with five classes, with the first label as the worst and the last one as the best.
LinearRegression.py -> In this file is implemented the linear regression, approsimating the result to the closer class label.
timeSeriesForecast.py -> In this file is implemented the forecast process with the arima model.
forecast.py -> In this file is there is the script for estimating the parameters of the arima model.
We did not provide the files with the data. If you want to try the entire code with also the files, please send an email to 93lorenzo@gmail.com