Application background Data & nbsp; within two months of 40000 customers 15 data field & nbsp; including customers & nbsp; & nbsp; & nbsp; & nbsp; & nbsp; a position, occupation, rank, annual income.. purchased products.Objective: to recommend a product that is suitable for the customer, and to recommend the purchase of the product.Data processing ; describe;  ,  ,  ,  ,  ,,  ,  ,  ,,  ,, -1,  ,  ,  ,  ,  ,  ,  ,,,  ,,, data normalization: eg: zip code similarity is not high according to the second tier line... The importance of dividing   Key Technology Feature selection random forest: the importance of judging features: thinking about why the importance of high   AHP; ;Model selection  : according to the amount of data, whether there i