Following on Sentimental Analysis, the aim is to aid Olist store with boosting sales and increase the revenue. There are multiple ways of building a recommender system. In order to select the model which should be used a lot of known algorithms are going to be trained on the dataset (using hyperparameter optimization for those that support it) and the best one is going to be picked.
To this part we evaluated 7 known recommendation algorithms to select the one who serves our needs best. SVDpp with the following parameters: n_factors=50, n_epochs=30, lr_all=0.01, reg_all=0.1 and SlopeOne were identified as best performing recommendation algorithms in studied case.