User Product Recommendation System Using KNN-Means and Singular Value Decomposition

In the course of our regular lives, each and every one of us frequently makes use of the e-commerce websites in order to purchase a variety of goods. These results in a rising diversity of consumer demand, which makes it difficult for a retail company to deliver the appropriate products in accordance to the tastes of its customers. These e-commerce websites make use of a variety of recommendation systems in order to give the consumer a satisfying experience when they are shopping online. A technique that may be used to overcome this obstacle is called a recommender system. With the help of product recommendation, it is possible to meet the requirements of customers, which assists in sustaining loyal clients while also attracting new customers. In this project, we propose a system for product based recommendation by using hybrid machine learning algorithms which consists of the combination of KNN-Means and Singular Value Decomposition for the purpose of improving the overall excellence of our rating and recommendation system, that improvises the disadvantages of collaborative filtering on their own. This method has the benefit that there is no necessity for the development of a brand new algorithm in order to calculate the forecasts. The trials demonstrate that our strategy can provide superior outcomes, and our integrated model provides an exceptionally high level of accurate predictions.