Let's say I have a dataset with pricing information for the same flight during the past year. So, for a flight departing on day D, I have the available price from D-130 up to D (departure day). Then the same for flights on the other 365 days of the year for the same departing time every day. Does this make sense? I want to see if today's price for any departure day in the future is higher or lower than expected, hence if it's expected to rise or not.
Plotting this data on a chart, the X axis are the days prior to departure, and the Y axis the price, I get the following:
How can I shape the data so I can train a model for price prediction? I can't see a clear trend. The price just below 90 means that it's the max price and it's not discounted.