What type of machine learning model should be used?
Classification month-to-month using supervised learning of the 200 categories based on claim contents.
Reinforcement learning using claim IDs and timestamps where the agent will identify how many claims in each category to expect from month to month.
Forecasting using claim IDs and timestamps to identify how many claims in each category to expect from month to month.
Classification with supervised learning of the categories for which partial information on claim contents is provided, and forecasting using claim IDs and timestamps for all other categories.
Explanations:
Classification is used for predicting categories, not the number of claims.
Reinforcement learning is not ideal for forecasting quantities like claim numbers over time.
Forecasting is the right approach to predict the number of claims in each category over time.
A mix of classification and forecasting is unnecessary and not ideal for this task.
I weigh that the answer is:
Forecasting using claim IDs and timestamps to identify how many claims in each category to expect from month to month.