Have anyone worked on Time Series Classification? I am working on a problem now where I have to forecast the change in the insurance carrier.
The inputs I have now so far are the Insurance Carrier Name, Companies, and Premium amount of each company.
So the goal is to build a model that takes these inputs and forecasts whether the company would change its carrier in the future or not.
Still, now I tried windowing techniques but I feel like I’m missing out somewhere. There are almost 99 different companies.
Any help or tidbits to get started on this would be highly appreciated.
This is how the sample data looks.