I've been working on the Netflix project. Given ratings that customers have given certain movies, we have been assigned the task of predicting the ratings they would give to other movies. We have been using two caches - one with the average ratings for each movie and one of the average ratings for each user. The target is to get the RMSE to be less than 1.0. So far, we have gotten the RMSE to be 1.003 but we think that we will need to use another cache to get lower results. That said, we have been created a cache with average ratings of all movies by year but so far, that has not shown to be better.What's in your way?
In preparation for interviews and career fair next week, I am brushing up on algorithms, past projects, and machine learning. I've had major breakthroughs in my robotics project this week and hopefully, this trend will continue. I have a large number of projects that I would like to be working on that I can't. However, I have now partnered with a few other people and progress is much better. It would be nice to get some of these on GitHub.What will you do next week?
I'm going to first get Netflix to run with an RMSE lower than 1.0. We will need to get it to pass all the requirements. I think that will be done in the next two days. On Wednesday, I'll be off to get my citizenship. I'll probably have to miss class for that but hopefully, I will be back in Austin Wednesday afternoon. I have two interviews scheduled for next week and of course, career fair is on Thursday.What’s my experience of the class?
I like the Netflix project. I was a little confused after reading the description but Prof. Downing explained the assignment in class and it is not too bad.What's my pick-of-the-week or tip-of-the-week?
Trump2Cash is a funny project that is currently trending on GitHub. While I would not necessarily recommend using it, you should check it out! It analyzes the President's tweets using sentiment analysis and automatically trades stocks when he mentions a publicly traded company.