The well-known grocery chain "Tosco & Spency" would like to differentiate its offering and marketing strategy based on two distinct consumer groups. Indeed, there are two types of clients who go to the supermarket: the first are more likely to buy a few high-priced items, while the later are more likely to buy a large number of low-priced items. They give you a training set of historical data with client characteristics and a label indicating whether the consumer is in the first class.Decision tree classifier, Random forest classifier and naive bayes classifier techniques are used to forecast whether a new consumer will buy a few expensive products rather than a large number of cheap ones.The performance of the model is evaluated using a hold out test set On comparing all the three classification techniques, Random Forest classifier has given better accuracy and is consequently used to make predictions on the new customers. Another famous supermarket chain gives you a training set of historical data that includes information about each consumer as well as a numerical value reflecting the amount of money they spent in May 2021.They want you to anticipate not just if the consumer is likely to buy a few high-priced items, but also how much money they spend on average each month.In this case, the expenditure of a new customer is predicted using using three regression algorithms, linear regressor, random forest regressor and xgboost regressor. When compared to Random forest regressor and Xgboot regressor, Linear regressor has better mean squared error and therefore is used to make predictions on the unseen data set.