CLV forecasting through the use of RFM variables. This is the code I used for my Bachelor thesis. There are many Python scripts I have used to filter and clean the data as well as to create new variables and build some prediction modles. The comments are in italian. There are also the parquet files I used to dived my work into smaller parts and make it more accessible but i can't upload them since they are too big but they are easly obtainable running the script with the original data from the Kaggle challange https://www.kaggle.com/c/acquire-valued-shoppers-challenge.