I have generated the following semi-continuous sales data, where product x is not really having a lot of sales in the first and last three months of the year. The rest of the year, sales are clearly higher.
How can I derive a method using machine learning where the input is a point in time (probably the month 1 - 12) and the output is the chance (0 to 100%) the product will perform? Can such a relation be derived and will it work for data which doesn't have such strong seasonality? Clearly this is not linear regression but there is a certain relation between two variables (month, sales).
Ideally, I wish to iterate over a set of n products and calculate which one is most ideal (according to said chance) to sell at a specific point in time. Is this the correct way to tackle the problem?