See details of the challenge on www.NetflixPrize.com.
I registered in 2006 Dec. My initial algorithm was mostly based on creating customer categories based on their rating habits, seasonal moods etc. But I couldn't do better than 0.98. The problem was the customers that had refused to fit in any pattern. Their standard deviation of ratings in any group was pretty high (1.2+).
I gave up.
Last week I came up with an idea to take care of these unpredictable customers by linking IMDB genre data. The idea is to categorize the customers based on what kind of movies they like.
And I beat the Netflix's Cinematch 0.03 points. My new RMSE is 0.9287.
Now I am stuck again. Not sure what else to try.
3 comments:
Hey are you a professional journalist? This article is very well written, as compared to most other blogs i saw today….
anyhow thanks for the good read!
Your blog keeps getting better and better! Your older articles are not as good as newer ones you have a lot more creativity and originality now keep it up!
Good post and this mail helped me alot in my college assignement. Thanks you as your information.
Post a Comment