Challenge problem In our context, a music recommendation system need to be developed to predict the preferences of the existing users based on their listening histories. The whole report will be divided into 8 parts: (1) Introduction: present information of the whole project (2) Evaluation metric: present the evaluation metric used in this project to evaluate the performance of different models (3) Baseline method: the simplest method which provides an baseline for our evaluation (4) Challenge problem: the task we are about to deal with (5) Memorybased Collaborative filtering: user-based method and item-based method are introduced (6) Model-based Collaborative filtering: SVD method is introduced (7) Experiments: the experimental setup and the final results (8) Conclusion: the conclusion of the project. By relying on the Million Song Dataset, the data for the competition is completely open: almost everything is known and possibly available. We focus on collaborative filtering method in this project since it is the most prevalent and effective method final-report in the recommendation system literature. The Million Song Dataset Challenge aims at being the best possible offline evaluation of a music recommendation system. There are also recommender systems for experts, collaborators, jokes, restaurants, garments, financial services, life insurance, romantic partners (online dating), and Twitter pages. Recommender systems have become increasingly popular in recent years, and are utilized in a variety of areas including movies, music, news, books, research articles, search queries, social tags, and products in general. Recommender systems or recommendation systems are a subclass of information filtering system that seek to predict the rating or preference that a user would give to an item. Under this context, we choose to build a system that can automatically recommend new songs to clients based on their listening histories in the past. Whether it is which movie to watch at Friday night or is there any interesting new products available on the Amazon. Introduction During everyday life, people always need some advices when making the decisions. At last we concluded the project and propose some prospectives.
Then the experimental results are presented and discussed. In this report, we present the problem we plan to solve and the theoretic explanation of different algorithms.and also the problems we meet. At the same time, different algorithms of recommendation are compared based on Million Song Dataset Challenge.
1 Final report - Advanced Machine Learning project Million Song Dataset Challenge Xiaoxiao CHEN Yuxiang WANG Honglin LI Abstract The purpose of this project is to learn the basic principles of recommendation system.
0 kommentar(er)
