Project Blog

1. Make a live recommendation table and an old table to not mess up the live table.

2. Next time discuss how to incorporate natural words in the recommendation.

1. Create a Media Job in Yioop for the recommendation system so that it runs periodically.

2. Incorporate the timestamp feature so that it takes into account the changes since the last time recommendation system was run.

3. Updates the recommendation table accordingly.

1. Discuss integrating baseline predictor class.

2. Build a new table Recommendation with columns user'_id, recommend_id, rating, recommend_type, etc.

3. Make use of ITEM_SUMMARY table to build the recommendation system for Yioop

1. Start Deliverable 1. Building standalone baseline predictor class

1. CS 298 Project Proposal and discuss the deliverable

1. Discussed the deliverables and CS 298 project proposal

1. Review draft 2 - CS 297 Report

2. Uploading all the deliverables and report

1. Deliverable 4 code review

2. Review draft 1 - CS 297 Report

1. Discussed the implementation for deliverable 4 - Latent matrix factorisation using gradient descent

2. Discussed how to make the final report for CS 297

1. Presentation - HITS algorithm

2. Finished Deliverable 3 - Patch 2

1. Presentation - SVD for recommendations

2. Finished Deliverable 3 - Patch 1

3. Start Deliverable 3 - Patch 2

1. Presentation - Latent Matrix Factorization using gradient descent

2. Deliverable 3 - Yioop patch 1 progress update

1. Presentation - Practical Machine Learning Book by Ted Dunning and Ellen Friedman

2. Discuss the progress on Yioop Manage account Page

1. Reviewed the code.

2. Validate email address at the client side along with server side.

3. Password strength checks and form should not submit if all fields are not filled.

4. In manage account page add a join date field and filter based on join dates.

1. Reviewed the code and discussed about optimizing the code so that the computer doesn't crash while processing and analyzing 2.7 million yelp review records.

2. Make changes to the code and upload Deliverable 2

3. Present Practical Machine Learning Book by Ted Dunning and Ellen Friedman.

4. Find a list of stuff about Yioop user interface which I do not like.

5. Understand the Yioop Framework.

1. Discussed about baseline predictors implementation and predicting users rating for Yelp business/restaurants based on reviews.

2. Completed presenting Chapter 2 from who's #1 ? The science of rating and ranking

1. Present chapter 1 and 2 from who's #1 ? The science of rating and ranking

2. Deliverable 1 completion

3. Start Collaborative filtering Implementation

1. Presented the Max Flow Algorithms

2. Demonstrated the max flow implementation - Ford Fulkerson and push relabel algorithms

3. For next week present the Netflix recommender system

4. For next week present chapter 1 and 2 from who's #1 ? The science of rating and ranking

5. Finish the push relabel code and add simple test cases for the max flow implementation. Write a summary for the code uploaded.

1. Presented the Max Flow Algorithms

2. Discussed Networked Life Chapter 4

3. For next week demonstrate the Max Flow Algorithm - Deliverable 1 due

1. Prepare Presentation for Max Flow propblem. Implement Max Flow problem in PHP. Present the NETFLIX recommendation Engine algorithm - chapter 4 Networked Life.

2. Update Proposal

3. Finalize the two algorithms for recommendation engine which needs to be implemented.

1. Deliverable 1: Max Flow Presentation and Implementation.
2. Deliverable 2: Recommendation Engine Algorithms - Literature Review
3. Deliverable 3: Download Yioop. Suggest something on manage account page. Try to improve some aspect of manage group activity.
4. Deliverable 4: Recommendation Engine Algorithm Implementation.
5. Deliverable 5: CS297 Final Report
6. For next week: Read chapter 26 from Algorithms book