lets say I have historical data for prices and some additional information like article, location and maybe text like "higher price in this area due to less competition". Important is, that prices are made by experts who exactly know the town and insider infos like competitors prices and so on. A lot of data rows but only a few features, like:
123/london road/item x/text abc
456/new york /item y/null
Is something like this little information useful? Ideas that comes to my mind: -Regression to predict price (additional info: how prices rise the last years) -Clustering (Maybe you find price cluster due to the locations) -Use the magic of neuronal networks predict prices with this technic and use the additional input for sales people and let the neuronal network learn also from gathering this new data.
But I am not sure if a use case with such few features makes sense, maybe someone else had the same Problems and maybe find a way to add additional features I did not think due to similar use cases.
