Comparison of Different Recommendation Models based on Amazon Platform Dataset

As the society enters the era of” information explosion”, the problem of information overload has become more and more prominent, and different recommendation systems are born. However, it is particularly important to choose the right recommendation model because different recommendation models have different accuracy for binary classification problems. This paper mainly compares the Amazon platform dataset by using logistic regression, decision tree, random forest, and deep crossing models and their mixed models, and finally measuring their performance. The results show that the combination model by using stacking method with a deep crossing weight of 0.70, logistic regression weight of 0.10, and random forest weight of 0.20 achieved the highest accuracy of 0.7567.