I have a questions about the data that goes into the machine learning models.
I am trying to predict loan prepayment. I have static predictors such as borrower attributes, as well as predictors in time series, for example, mortgage rate of each month. I wonder how the data should look like for the model. One loan per record (using feature extraction to convert time series data to static) or one loan with multiple records. If it is one loan with multiple records, i want to know how the model will understand the information. I have attached example below.
Thank you for your help!