{ "id": 363624, "title": "Andrew Ng Recommender Systems", "url": "/competitions/otto-recommender-system/discussion/363624", "author_name": "", "post_date": "2022-11-02T13:53:20.720426400Z", "votes": 62, "comment_count": 15, "views": 0, "content": "

Hi all!

\n

I've asked a good friend to send me a few articles on Reccommender Systems to help me get a base before starting, so I am sharing these videos by Andrew Ng with you too: https://www.youtube.com/watch?v=giIXNoiqO_U&list=PLLssT5z_DsK-h9vYZkQkYNWcItqhlRJLN&index=97

\n

They are quite short and straight forward: from lecture 16.1 to 16.6

\n

Good luck!

", "messages": [ { "id": "2014315", "postDate": "11/02/2022 13:53:20", "content": "

Hi all!

\n

I've asked a good friend to send me a few articles on Reccommender Systems to help me get a base before starting, so I am sharing these videos by Andrew Ng with you too: https://www.youtube.com/watch?v=giIXNoiqO_U&list=PLLssT5z_DsK-h9vYZkQkYNWcItqhlRJLN&index=97

\n

They are quite short and straight forward: from lecture 16.1 to 16.6

\n

Good luck!

", "rawMarkdown": "Hi all!\n\nI've asked a good friend to send me a few articles on Reccommender Systems to help me get a base before starting, so I am sharing these videos by Andrew Ng with you too: https://www.youtube.com/watch?v=giIXNoiqO_U&list=PLLssT5z_DsK-h9vYZkQkYNWcItqhlRJLN&index=97\n\nThey are quite short and straight forward: from lecture 16.1 to 16.6\n\nGood luck!", "votes": null }, { "id": "2014558", "postDate": "11/02/2022 16:37:23", "content": "

Thanks for sharing @andradaolteanu πŸ˜„

", "rawMarkdown": "Thanks for sharing @andradaolteanu πŸ˜„", "votes": null }, { "id": "2014849", "postDate": "11/02/2022 22:50:19", "content": "

Legend πŸ˜„

\n

\"\"

", "rawMarkdown": "Legend πŸ˜„\n\n![](https://raw.githubusercontent.com/snnclsr/kaggle_images/main/andrew_ng.jpeg)", "votes": null }, { "id": "2014942", "postDate": "11/03/2022 01:47:08", "content": "

Thanks for sharing @andradaolteanu, it help a lot to new bees like me. Keep sharing and rocking.

", "rawMarkdown": "Thanks for sharing @andradaolteanu, it help a lot to new bees like me. Keep sharing and rocking.", "votes": null }, { "id": "2016142", "postDate": "11/03/2022 20:06:06", "content": "

Great resources! Thanks for the post @andradaolteanu

", "rawMarkdown": "Great resources! Thanks for the post @andradaolteanu", "votes": null }, { "id": "2016158", "postDate": "11/03/2022 20:18:25", "content": "

I loved his original course. It actually got updated: https://www.coursera.org/learn/unsupervised-learning-recommenders-reinforcement-learning?specialization=machine-learning-introduction I highly recommend it. (Both the old and new).
\nAndrew has helped me a great deal on my journey.

", "rawMarkdown": "I loved his original course. It actually got updated: https://www.coursera.org/learn/unsupervised-learning-recommenders-reinforcement-learning?specialization=machine-learning-introduction I highly recommend it. (Both the old and new). \nAndrew has helped me a great deal on my journey.", "votes": null }, { "id": "2016484", "postDate": "11/04/2022 04:09:04", "content": "

It is helpful.

", "rawMarkdown": "It is helpful.", "votes": null }, { "id": "2019675", "postDate": "11/06/2022 20:32:36", "content": "

Andrew in the new Electricity Generator, hats off, truly.

", "rawMarkdown": "Andrew in the new Electricity Generator, hats off, truly.", "votes": null }, { "id": "2020463", "postDate": "11/07/2022 14:01:10", "content": "

Thank You ^_^

", "rawMarkdown": "Thank You ^_^", "votes": null }, { "id": "2021439", "postDate": "11/08/2022 08:43:54", "content": "

Thanks for the resources. @andradaolteanu πŸ™Œ

", "rawMarkdown": "Thanks for the resources. @andradaolteanu πŸ™Œ", "votes": null }, { "id": "2021900", "postDate": "11/08/2022 14:45:25", "content": "

I have done the notes on the new and old videos of Andrew Ng on recommendation system, hope it is helpful to others.

\n

Lecture 16.2 β€” Recommender Systems Intro
\n00:00 - How important recommendation system to the real world?
\n01:30 - Why should we study recommendation system? (Did or do academics pay much attention to it? Can it teach us the power of learning features of ML?)
\n03:00 - Andrew presented the recommendation system problem with a simplest example

\n

Lecture 16.2 β€” Recommender Systems | Content Based Recommendations
\n00:00 - recap of the problem of recommendation system
\n00:23 - What is content-based recommendation system? How to describe the recommendation system problem in details with symbols (mathematically)?
\nHow To predict the rate of any user to any movie, we just need to find out the features of the user and the features of the movie; we need to use the existing ratings of movies by current users to find out the features of current users and current movies
\n06:12 - How to learn the features of a single user by designing a simple loss function?
\n10:06 - How to learn the features of all users by combining all of the losses based on the loss function above?
\n11:25 - How to calculate the derivaties or doing gradient descent on user features?
\n13:37 - Why called content based? using features of movies (contents) to find out features of users and then to make predictions on ratings

\n

Lecture 16.3 β€” Recommender Systems | Collaborative filtering Recommendations Intro
\n00:00 - What's special about collaborative filtering compared with content-based recommendation system? It can learn all the features itself without giving features of movies first like content-based approach
\n00:21 - How to present collaborative filtering in terms of knowing user features and using rating data to find out features of movies and then make predictions on ratings (well, it is a just another content based approach using user features instead of movie features)
\n04:39 - What is the simplest loss function to help do gradient descent to find out features of all movies
\n06:56 - Why called collaborative filtering? what are collaborating? (between two content based approaches) why called filtering? What ensures such collaboration can learning possible? (data is good: a movie is rated by multiple users, a user has rated multiple movies)
\n09:15 - Andrew Ng explained why called collaboration filtering

\n

Lecture 16.4 β€” Recommender Systems | Collaborative Filtering Algorithm
\n00:00 - How to combine the loss function of both content-based approaches to do gradient descent on both features of users and movies, instead of finding features of users and then movies sequentially?
\n04:40 - Andrew present the collaborative algorithm in steps

\n

Lecture 16.5 β€” Recommender Systems | Vectorization Low Rank Matrix Factorization
\n00:00 - What is low-rank matrix factorization? put all features of all movies into a matrix, all features of all users into a matrix and put all rating prediction as a dot product of two matricies
\n05:07 - How can collaborative filtering enable recommendations of similar movies to a user? As it finds out all the features of all movies, we can compare the similarity of movie feature vectors to find out similar movies

\n

Lecture 16.6 β€” Recommender Systems | Implementational Detail Mean Normalization
\n00:00 - How a new user without giving any rating create challenges on learning based on the loss of our recommendation system algo learnt above
\n03:23 - How mean normalization on all movies' ratings on the rows (movie on row, user on col) come to rescue? Why it makes sense? (assume the new user will give the average ratings to all movies)
\n07:29 - What about having movies without any user giving ratings? What problem does taking mean normalization on ratings on col? (mean normalization on a user's ratings of all movies does not make much sense )

\n

Below are the actual new videos from this new course

\n

Binary labels: favs, likes and clicks
\n00:00 - What is the binary label problem application in recsys?
\n01:50 - How to define label 1 and label 0?
\n04:04 - How to move from regression to binary prediction in recsys?
\n05:16 - How to modify loss function for binary labels?

\n

Finding related items
\n00:00 - how to find related items using features of movies or products
\n03:24 - what are the limitations of collaborative filter? (cold start problem and not-using side information problem)

\n

Collaborative filtering vs Content-based filtering
\n00:00 - How collaboratibe filtering differ from content based filtering? (content-based filtering use side info on both users and items)
\n01:57 - What do the features or side informations of users or items look like? What's the problem to solve in order to utilize the features?
\n05:44 - How does the features of users and items can help make predictions of ratings? The key is to convert the side info of users and items into the actual features (same length) of users and items in making predictions
\n09:10 - summary

\n

Content-based Filtering | Deep learning for content-based filtering
\n00:00 - How deep learning (two neuralnets) help to convert side info into features (same length) of users and items?
\n03:07 - How to do the above with a single neuralnet?
\n06:02 - How to find similar items
\n08:09 - Sum up: deep learning can easily merge two neuralnets together; how to in practice make the computation of large amount of side info of users and items into actual features fast

\n

9.10 Advanced implementation | Recommending from a large catalogue
\n00:00 - Why it is a challenge to make predictions when there are millions of users and items?
\n01:17 - How to make use of features of users and items to recommend for existing user daily? Retrieval of related items for the user
\n04:03 - How to use the model to predict on the retrieved items in order to speed up or ease the computation? Rank by using the neuralnet model
\n05:26 - How to decide how many items to select in retrieval step?
\n07:09 - sum up

\n

9.11 Advanced implementation | Ethical use of recommender systems
\n00:00 - goals of recommendation system? what are the suspicous ones?
\n04:28 - Ad examples
\n07:51 - other problematic cases

", "rawMarkdown": "I have done the notes on the new and old videos of Andrew Ng on recommendation system, hope it is helpful to others. \n\n\n**Lecture 16.2 β€” Recommender Systems Intro**\n[00:00](https://youtu.be/giIXNoiqO_U?list=PLLssT5z_DsK-h9vYZkQkYNWcItqhlRJLN&t=0) - How important recommendation system to the real world?\n[01:30](https://youtu.be/giIXNoiqO_U?list=PLLssT5z_DsK-h9vYZkQkYNWcItqhlRJLN&t=90) - Why should we study recommendation system? (Did or do academics pay much attention to it? Can it teach us the power of learning features of ML?)\n[03:00](https://youtu.be/giIXNoiqO_U?list=PLLssT5z_DsK-h9vYZkQkYNWcItqhlRJLN&t=180) - Andrew presented the recommendation system problem with a simplest example\n\n**Lecture 16.2 β€” Recommender Systems | Content Based Recommendations**\n[00:00](https://youtu.be/9siFuMMHNIA?list=PLLssT5z_DsK-h9vYZkQkYNWcItqhlRJLN&t=0) - recap of the problem of recommendation system \n[00:23](https://youtu.be/9siFuMMHNIA?list=PLLssT5z_DsK-h9vYZkQkYNWcItqhlRJLN&t=23) - What is content-based recommendation system? How to describe the recommendation system problem in details with symbols (mathematically)? \nHow To predict the rate of any user to any movie, we just need to find out the features of the user and the features of the movie; we need to use the existing ratings of movies by current users to find out the features of current users and current movies\n[06:12](https://youtu.be/9siFuMMHNIA?list=PLLssT5z_DsK-h9vYZkQkYNWcItqhlRJLN&t=372) - How to learn the features of a single user by designing a simple loss function?\n[10:06](https://youtu.be/9siFuMMHNIA?list=PLLssT5z_DsK-h9vYZkQkYNWcItqhlRJLN&t=606) - How to learn the features of all users by combining all of the losses based on the loss function above?\n[11:25](https://youtu.be/9siFuMMHNIA?list=PLLssT5z_DsK-h9vYZkQkYNWcItqhlRJLN&t=685) - How to calculate the derivaties or doing gradient descent on user features?\n[13:37](https://youtu.be/9siFuMMHNIA?list=PLLssT5z_DsK-h9vYZkQkYNWcItqhlRJLN&t=817) - Why called content based? using features of movies (contents) to find out features of users and then to make predictions on ratings\n\n**Lecture 16.3 β€” Recommender Systems | Collaborative filtering Recommendations Intro**\n[00:00](https://youtu.be/9AP-DgFBNP4?list=PLLssT5z_DsK-h9vYZkQkYNWcItqhlRJLN&t=0) - What's special about collaborative filtering compared with content-based recommendation system? It can learn all the features itself without giving features of movies first like content-based approach\n[00:21](https://youtu.be/9AP-DgFBNP4?list=PLLssT5z_DsK-h9vYZkQkYNWcItqhlRJLN&t=21) - How to present collaborative filtering in terms of knowing user features and using rating data to find out features of movies and then make predictions on ratings (well, it is a just another content based approach using user features instead of movie features)\n[04:39](https://youtu.be/9AP-DgFBNP4?list=PLLssT5z_DsK-h9vYZkQkYNWcItqhlRJLN&t=279) - What is the simplest loss function to help do gradient descent to find out features of all movies\n[06:56](https://youtu.be/9AP-DgFBNP4?list=PLLssT5z_DsK-h9vYZkQkYNWcItqhlRJLN&t=416) - Why called collaborative filtering? what are collaborating? (between two content based approaches) why called filtering? What ensures such collaboration can learning possible? (data is good: a movie is rated by multiple users, a user has rated multiple movies)\n[09:15](https://youtu.be/9AP-DgFBNP4?list=PLLssT5z_DsK-h9vYZkQkYNWcItqhlRJLN&t=555) - Andrew Ng explained why called collaboration filtering\n\n\n**Lecture 16.4 β€” Recommender Systems | Collaborative Filtering Algorithm**\n[00:00](https://youtu.be/YW2b8La2ICo?list=PLLssT5z_DsK-h9vYZkQkYNWcItqhlRJLN&t=0) - How to combine the loss function of both content-based approaches to do gradient descent on both features of users and movies, instead of finding features of users and then movies sequentially?\n[04:40](https://youtu.be/YW2b8La2ICo?list=PLLssT5z_DsK-h9vYZkQkYNWcItqhlRJLN&t=355) - Andrew present the collaborative algorithm in steps\n\n**Lecture 16.5 β€” Recommender Systems | Vectorization Low Rank Matrix Factorization**\n[00:00](https://youtu.be/5R1xOJOFRzs?list=PLLssT5z_DsK-h9vYZkQkYNWcItqhlRJLN&t=0) - What is low-rank matrix factorization? put all features of all movies into a matrix, all features of all users into a matrix and put all rating prediction as a dot product of two matricies\n[05:07](https://youtu.be/5R1xOJOFRzs?list=PLLssT5z_DsK-h9vYZkQkYNWcItqhlRJLN&t=307) - How can collaborative filtering enable recommendations of similar movies to a user? As it finds out all the features of all movies, we can compare the similarity of movie feature vectors to find out similar movies\n\n**Lecture 16.6 β€” Recommender Systems | Implementational Detail Mean Normalization**\n[00:00](https://youtu.be/Am9fhp2Q91o?list=PLLssT5z_DsK-h9vYZkQkYNWcItqhlRJLN&t=0) - How a new user without giving any rating create challenges on learning based on the loss of our recommendation system algo learnt above\n[03:23](https://youtu.be/Am9fhp2Q91o?list=PLLssT5z_DsK-h9vYZkQkYNWcItqhlRJLN&t=203) - How mean normalization on all movies' ratings on the rows (movie on row, user on col) come to rescue? Why it makes sense? (assume the new user will give the average ratings to all movies)\n[07:29](https://youtu.be/Am9fhp2Q91o?list=PLLssT5z_DsK-h9vYZkQkYNWcItqhlRJLN&t=449) - What about having movies without any user giving ratings? What problem does taking mean normalization on ratings on col? (mean normalization on a user's ratings of all movies does not make much sense )\n\nBelow are the actual new videos from this new course\n\n**Binary labels: favs, likes and clicks** \n[00:00](https://youtu.be/tnIiuLQk63I?t=0) - What is the binary label problem application in recsys?\n[01:50](https://youtu.be/tnIiuLQk63I?t=110) - How to define label 1 and label 0?\n[04:04](https://youtu.be/tnIiuLQk63I?t=244) - How to move from regression to binary prediction in recsys? \n[05:16](https://youtu.be/tnIiuLQk63I?t=316) - How to modify loss function for binary labels?\n\n**Finding related items**\n[00:00](https://youtu.be/uXMa7YwDVbE?t=0) - how to find related items using features of movies or products\n03:24 - what are the limitations of collaborative filter? (cold start problem and not-using side information problem)\n\n**Collaborative filtering vs Content-based filtering** \n[00:00](https://youtu.be/UgSBaD7s5HU?t=0) - How collaboratibe filtering differ from content based filtering? (content-based filtering use side info on both users and items) \n[01:57](https://youtu.be/UgSBaD7s5HU?t=116) - What do the features or side informations of users or items look like? What's the problem to solve in order to utilize the features?\n[05:44](https://youtu.be/UgSBaD7s5HU?t=345) - How does the features of users and items can help make predictions of ratings? The key is to convert the side info of users and items into the actual features (same length) of users and items in making predictions \n[09:10](https://youtu.be/UgSBaD7s5HU?t=550) - summary\n\n**Content-based Filtering | Deep learning for content-based filtering**\n[00:00](https://youtu.be/jhgnQB7fYKM?t=0) - How deep learning (two neuralnets) help to convert side info into features (same length) of users and items?\n[03:07](https://youtu.be/jhgnQB7fYKM?t=187) - How to do the above with a single neuralnet?\n[06:02](https://youtu.be/jhgnQB7fYKM?t=362) - How to find similar items \n[08:09](https://youtu.be/jhgnQB7fYKM?t=489) - Sum up: deep learning can easily merge two neuralnets together; how to in practice make the computation of large amount of side info of users and items into actual features fast\n\n**9.10 Advanced implementation | Recommending from a large catalogue**\n[00:00](https://www.youtube.com/watch?v=nNJPU5fwc8E&list=PLxfEOJXRm7eZKJyovNH-lE3ooXTsOCvfC&index=123&loop=0) - Why it is a challenge to make predictions when there are millions of users and items?\n[01:17](https://youtu.be/nNJPU5fwc8E?list=PLxfEOJXRm7eZKJyovNH-lE3ooXTsOCvfC&t=77) - How to make use of features of users and items to recommend for existing user daily? Retrieval of related items for the user\n[04:03](https://youtu.be/nNJPU5fwc8E?list=PLxfEOJXRm7eZKJyovNH-lE3ooXTsOCvfC&t=243) - How to use the model to predict on the retrieved items in order to speed up or ease the computation? Rank by using the neuralnet model\n[05:26](https://youtu.be/nNJPU5fwc8E?list=PLxfEOJXRm7eZKJyovNH-lE3ooXTsOCvfC&t=326) - How to decide how many items to select in retrieval step?\n[07:09](https://youtu.be/nNJPU5fwc8E?list=PLxfEOJXRm7eZKJyovNH-lE3ooXTsOCvfC&t=429) - sum up\n\n**9.11 Advanced implementation | Ethical use of recommender systems**\n[00:00](https://youtu.be/GGnPWaNjaGM?list=PLxfEOJXRm7eZKJyovNH-lE3ooXTsOCvfC) - goals of recommendation system? what are the suspicous ones?\n[04:28](https://youtu.be/GGnPWaNjaGM?list=PLxfEOJXRm7eZKJyovNH-lE3ooXTsOCvfC&t=268) - Ad examples\n[07:51](https://youtu.be/GGnPWaNjaGM?list=PLxfEOJXRm7eZKJyovNH-lE3ooXTsOCvfC&t=471) - other problematic cases", "votes": null }, { "id": "2021972", "postDate": "11/08/2022 15:46:08", "content": "

thank you for sharing

", "rawMarkdown": "thank you for sharing", "votes": null }, { "id": "2022448", "postDate": "11/09/2022 03:44:31", "content": "

You are welcome

", "rawMarkdown": "You are welcome", "votes": null }, { "id": "2022796", "postDate": "11/09/2022 09:46:10", "content": "

It is helpful, thanks a lot.

", "rawMarkdown": "It is helpful, thanks a lot.", "votes": null }, { "id": "2026389", "postDate": "11/12/2022 03:04:36", "content": "

Outstanding remark… Thanks a lot!

", "rawMarkdown": "Outstanding remark... Thanks a lot!", "votes": null }, { "id": "2051806", "postDate": "12/01/2022 16:20:47", "content": "

Thank you very much for sharing!

", "rawMarkdown": "Thank you very much for sharing!", "votes": null } ], "comments": [ { "id": 2014558, "author_name": "mrgabrielblins", "author_url": "", "post_date": "11/02/2022 16:37:23", "content": "

Thanks for sharing @andradaolteanu πŸ˜„

", "votes": null, "replies": [] }, { "id": 2014849, "author_name": "snnclsr", "author_url": "", "post_date": "11/02/2022 22:50:19", "content": "

Legend πŸ˜„

\n

\"\"

", "votes": null, "replies": [] }, { "id": 2014942, "author_name": "rahul029", "author_url": "", "post_date": "11/03/2022 01:47:08", "content": "

Thanks for sharing @andradaolteanu, it help a lot to new bees like me. Keep sharing and rocking.

", "votes": null, "replies": [] }, { "id": 2016142, "author_name": "ravi20076", "author_url": "", "post_date": "11/03/2022 20:06:06", "content": "

Great resources! Thanks for the post @andradaolteanu

", "votes": null, "replies": [] }, { "id": 2016158, "author_name": "ollibolli", "author_url": "", "post_date": "11/03/2022 20:18:25", "content": "

I loved his original course. It actually got updated: https://www.coursera.org/learn/unsupervised-learning-recommenders-reinforcement-learning?specialization=machine-learning-introduction I highly recommend it. (Both the old and new).
\nAndrew has helped me a great deal on my journey.

", "votes": null, "replies": [] }, { "id": 2016484, "author_name": "hasanrahim123", "author_url": "", "post_date": "11/04/2022 04:09:04", "content": "

It is helpful.

", "votes": null, "replies": [] }, { "id": 2019675, "author_name": "priyanshuprajapatis", "author_url": "", "post_date": "11/06/2022 20:32:36", "content": "

Andrew in the new Electricity Generator, hats off, truly.

", "votes": null, "replies": [] }, { "id": 2020463, "author_name": "javedahmedshanto", "author_url": "", "post_date": "11/07/2022 14:01:10", "content": "

Thank You ^_^

", "votes": null, "replies": [] }, { "id": 2021439, "author_name": "michaelswaminathan", "author_url": "", "post_date": "11/08/2022 08:43:54", "content": "

Thanks for the resources. @andradaolteanu πŸ™Œ

", "votes": null, "replies": [] }, { "id": 2021900, "author_name": "danielliao", "author_url": "", "post_date": "11/08/2022 14:45:25", "content": "

I have done the notes on the new and old videos of Andrew Ng on recommendation system, hope it is helpful to others.

\n

Lecture 16.2 β€” Recommender Systems Intro
\n00:00 - How important recommendation system to the real world?
\n01:30 - Why should we study recommendation system? (Did or do academics pay much attention to it? Can it teach us the power of learning features of ML?)
\n03:00 - Andrew presented the recommendation system problem with a simplest example

\n

Lecture 16.2 β€” Recommender Systems | Content Based Recommendations
\n00:00 - recap of the problem of recommendation system
\n00:23 - What is content-based recommendation system? How to describe the recommendation system problem in details with symbols (mathematically)?
\nHow To predict the rate of any user to any movie, we just need to find out the features of the user and the features of the movie; we need to use the existing ratings of movies by current users to find out the features of current users and current movies
\n06:12 - How to learn the features of a single user by designing a simple loss function?
\n10:06 - How to learn the features of all users by combining all of the losses based on the loss function above?
\n11:25 - How to calculate the derivaties or doing gradient descent on user features?
\n13:37 - Why called content based? using features of movies (contents) to find out features of users and then to make predictions on ratings

\n

Lecture 16.3 β€” Recommender Systems | Collaborative filtering Recommendations Intro
\n00:00 - What's special about collaborative filtering compared with content-based recommendation system? It can learn all the features itself without giving features of movies first like content-based approach
\n00:21 - How to present collaborative filtering in terms of knowing user features and using rating data to find out features of movies and then make predictions on ratings (well, it is a just another content based approach using user features instead of movie features)
\n04:39 - What is the simplest loss function to help do gradient descent to find out features of all movies
\n06:56 - Why called collaborative filtering? what are collaborating? (between two content based approaches) why called filtering? What ensures such collaboration can learning possible? (data is good: a movie is rated by multiple users, a user has rated multiple movies)
\n09:15 - Andrew Ng explained why called collaboration filtering

\n

Lecture 16.4 β€” Recommender Systems | Collaborative Filtering Algorithm
\n00:00 - How to combine the loss function of both content-based approaches to do gradient descent on both features of users and movies, instead of finding features of users and then movies sequentially?
\n04:40 - Andrew present the collaborative algorithm in steps

\n

Lecture 16.5 β€” Recommender Systems | Vectorization Low Rank Matrix Factorization
\n00:00 - What is low-rank matrix factorization? put all features of all movies into a matrix, all features of all users into a matrix and put all rating prediction as a dot product of two matricies
\n05:07 - How can collaborative filtering enable recommendations of similar movies to a user? As it finds out all the features of all movies, we can compare the similarity of movie feature vectors to find out similar movies

\n

Lecture 16.6 β€” Recommender Systems | Implementational Detail Mean Normalization
\n00:00 - How a new user without giving any rating create challenges on learning based on the loss of our recommendation system algo learnt above
\n03:23 - How mean normalization on all movies' ratings on the rows (movie on row, user on col) come to rescue? Why it makes sense? (assume the new user will give the average ratings to all movies)
\n07:29 - What about having movies without any user giving ratings? What problem does taking mean normalization on ratings on col? (mean normalization on a user's ratings of all movies does not make much sense )

\n

Below are the actual new videos from this new course

\n

Binary labels: favs, likes and clicks
\n00:00 - What is the binary label problem application in recsys?
\n01:50 - How to define label 1 and label 0?
\n04:04 - How to move from regression to binary prediction in recsys?
\n05:16 - How to modify loss function for binary labels?

\n

Finding related items
\n00:00 - how to find related items using features of movies or products
\n03:24 - what are the limitations of collaborative filter? (cold start problem and not-using side information problem)

\n

Collaborative filtering vs Content-based filtering
\n00:00 - How collaboratibe filtering differ from content based filtering? (content-based filtering use side info on both users and items)
\n01:57 - What do the features or side informations of users or items look like? What's the problem to solve in order to utilize the features?
\n05:44 - How does the features of users and items can help make predictions of ratings? The key is to convert the side info of users and items into the actual features (same length) of users and items in making predictions
\n09:10 - summary

\n

Content-based Filtering | Deep learning for content-based filtering
\n00:00 - How deep learning (two neuralnets) help to convert side info into features (same length) of users and items?
\n03:07 - How to do the above with a single neuralnet?
\n06:02 - How to find similar items
\n08:09 - Sum up: deep learning can easily merge two neuralnets together; how to in practice make the computation of large amount of side info of users and items into actual features fast

\n

9.10 Advanced implementation | Recommending from a large catalogue
\n00:00 - Why it is a challenge to make predictions when there are millions of users and items?
\n01:17 - How to make use of features of users and items to recommend for existing user daily? Retrieval of related items for the user
\n04:03 - How to use the model to predict on the retrieved items in order to speed up or ease the computation? Rank by using the neuralnet model
\n05:26 - How to decide how many items to select in retrieval step?
\n07:09 - sum up

\n

9.11 Advanced implementation | Ethical use of recommender systems
\n00:00 - goals of recommendation system? what are the suspicous ones?
\n04:28 - Ad examples
\n07:51 - other problematic cases

", "votes": null, "replies": [ { "id": 2021972, "author_name": "cdeotte", "author_url": "", "post_date": "11/08/2022 15:46:08", "content": "

thank you for sharing

", "votes": null, "replies": [] }, { "id": 2022448, "author_name": "danielliao", "author_url": "", "post_date": "11/09/2022 03:44:31", "content": "

You are welcome

", "votes": null, "replies": [] }, { "id": 2051806, "author_name": "leiwong", "author_url": "", "post_date": "12/01/2022 16:20:47", "content": "

Thank you very much for sharing!

", "votes": null, "replies": [] } ] }, { "id": 2022796, "author_name": "deanpp", "author_url": "", "post_date": "11/09/2022 09:46:10", "content": "

It is helpful, thanks a lot.

", "votes": null, "replies": [] }, { "id": 2026389, "author_name": "adrianfaela", "author_url": "", "post_date": "11/12/2022 03:04:36", "content": "

Outstanding remark… Thanks a lot!

", "votes": null, "replies": [] } ], "raw_markdown_by_id": { "2014315": "Hi all!\n\nI've asked a good friend to send me a few articles on Reccommender Systems to help me get a base before starting, so I am sharing these videos by Andrew Ng with you too: https://www.youtube.com/watch?v=giIXNoiqO_U&list=PLLssT5z_DsK-h9vYZkQkYNWcItqhlRJLN&index=97\n\nThey are quite short and straight forward: from lecture 16.1 to 16.6\n\nGood luck!", "2014558": "Thanks for sharing @andradaolteanu πŸ˜„", "2014849": "Legend πŸ˜„\n\n![](https://raw.githubusercontent.com/snnclsr/kaggle_images/main/andrew_ng.jpeg)", "2014942": "Thanks for sharing @andradaolteanu, it help a lot to new bees like me. Keep sharing and rocking.", "2016142": "Great resources! Thanks for the post @andradaolteanu", "2016158": "I loved his original course. It actually got updated: https://www.coursera.org/learn/unsupervised-learning-recommenders-reinforcement-learning?specialization=machine-learning-introduction I highly recommend it. (Both the old and new). \nAndrew has helped me a great deal on my journey.", "2016484": "It is helpful.", "2019675": "Andrew in the new Electricity Generator, hats off, truly.", "2020463": "Thank You ^_^", "2021439": "Thanks for the resources. @andradaolteanu πŸ™Œ", "2021900": "I have done the notes on the new and old videos of Andrew Ng on recommendation system, hope it is helpful to others. \n\n\n**Lecture 16.2 β€” Recommender Systems Intro**\n[00:00](https://youtu.be/giIXNoiqO_U?list=PLLssT5z_DsK-h9vYZkQkYNWcItqhlRJLN&t=0) - How important recommendation system to the real world?\n[01:30](https://youtu.be/giIXNoiqO_U?list=PLLssT5z_DsK-h9vYZkQkYNWcItqhlRJLN&t=90) - Why should we study recommendation system? (Did or do academics pay much attention to it? Can it teach us the power of learning features of ML?)\n[03:00](https://youtu.be/giIXNoiqO_U?list=PLLssT5z_DsK-h9vYZkQkYNWcItqhlRJLN&t=180) - Andrew presented the recommendation system problem with a simplest example\n\n**Lecture 16.2 β€” Recommender Systems | Content Based Recommendations**\n[00:00](https://youtu.be/9siFuMMHNIA?list=PLLssT5z_DsK-h9vYZkQkYNWcItqhlRJLN&t=0) - recap of the problem of recommendation system \n[00:23](https://youtu.be/9siFuMMHNIA?list=PLLssT5z_DsK-h9vYZkQkYNWcItqhlRJLN&t=23) - What is content-based recommendation system? How to describe the recommendation system problem in details with symbols (mathematically)? \nHow To predict the rate of any user to any movie, we just need to find out the features of the user and the features of the movie; we need to use the existing ratings of movies by current users to find out the features of current users and current movies\n[06:12](https://youtu.be/9siFuMMHNIA?list=PLLssT5z_DsK-h9vYZkQkYNWcItqhlRJLN&t=372) - How to learn the features of a single user by designing a simple loss function?\n[10:06](https://youtu.be/9siFuMMHNIA?list=PLLssT5z_DsK-h9vYZkQkYNWcItqhlRJLN&t=606) - How to learn the features of all users by combining all of the losses based on the loss function above?\n[11:25](https://youtu.be/9siFuMMHNIA?list=PLLssT5z_DsK-h9vYZkQkYNWcItqhlRJLN&t=685) - How to calculate the derivaties or doing gradient descent on user features?\n[13:37](https://youtu.be/9siFuMMHNIA?list=PLLssT5z_DsK-h9vYZkQkYNWcItqhlRJLN&t=817) - Why called content based? using features of movies (contents) to find out features of users and then to make predictions on ratings\n\n**Lecture 16.3 β€” Recommender Systems | Collaborative filtering Recommendations Intro**\n[00:00](https://youtu.be/9AP-DgFBNP4?list=PLLssT5z_DsK-h9vYZkQkYNWcItqhlRJLN&t=0) - What's special about collaborative filtering compared with content-based recommendation system? It can learn all the features itself without giving features of movies first like content-based approach\n[00:21](https://youtu.be/9AP-DgFBNP4?list=PLLssT5z_DsK-h9vYZkQkYNWcItqhlRJLN&t=21) - How to present collaborative filtering in terms of knowing user features and using rating data to find out features of movies and then make predictions on ratings (well, it is a just another content based approach using user features instead of movie features)\n[04:39](https://youtu.be/9AP-DgFBNP4?list=PLLssT5z_DsK-h9vYZkQkYNWcItqhlRJLN&t=279) - What is the simplest loss function to help do gradient descent to find out features of all movies\n[06:56](https://youtu.be/9AP-DgFBNP4?list=PLLssT5z_DsK-h9vYZkQkYNWcItqhlRJLN&t=416) - Why called collaborative filtering? what are collaborating? (between two content based approaches) why called filtering? What ensures such collaboration can learning possible? (data is good: a movie is rated by multiple users, a user has rated multiple movies)\n[09:15](https://youtu.be/9AP-DgFBNP4?list=PLLssT5z_DsK-h9vYZkQkYNWcItqhlRJLN&t=555) - Andrew Ng explained why called collaboration filtering\n\n\n**Lecture 16.4 β€” Recommender Systems | Collaborative Filtering Algorithm**\n[00:00](https://youtu.be/YW2b8La2ICo?list=PLLssT5z_DsK-h9vYZkQkYNWcItqhlRJLN&t=0) - How to combine the loss function of both content-based approaches to do gradient descent on both features of users and movies, instead of finding features of users and then movies sequentially?\n[04:40](https://youtu.be/YW2b8La2ICo?list=PLLssT5z_DsK-h9vYZkQkYNWcItqhlRJLN&t=355) - Andrew present the collaborative algorithm in steps\n\n**Lecture 16.5 β€” Recommender Systems | Vectorization Low Rank Matrix Factorization**\n[00:00](https://youtu.be/5R1xOJOFRzs?list=PLLssT5z_DsK-h9vYZkQkYNWcItqhlRJLN&t=0) - What is low-rank matrix factorization? put all features of all movies into a matrix, all features of all users into a matrix and put all rating prediction as a dot product of two matricies\n[05:07](https://youtu.be/5R1xOJOFRzs?list=PLLssT5z_DsK-h9vYZkQkYNWcItqhlRJLN&t=307) - How can collaborative filtering enable recommendations of similar movies to a user? As it finds out all the features of all movies, we can compare the similarity of movie feature vectors to find out similar movies\n\n**Lecture 16.6 β€” Recommender Systems | Implementational Detail Mean Normalization**\n[00:00](https://youtu.be/Am9fhp2Q91o?list=PLLssT5z_DsK-h9vYZkQkYNWcItqhlRJLN&t=0) - How a new user without giving any rating create challenges on learning based on the loss of our recommendation system algo learnt above\n[03:23](https://youtu.be/Am9fhp2Q91o?list=PLLssT5z_DsK-h9vYZkQkYNWcItqhlRJLN&t=203) - How mean normalization on all movies' ratings on the rows (movie on row, user on col) come to rescue? Why it makes sense? (assume the new user will give the average ratings to all movies)\n[07:29](https://youtu.be/Am9fhp2Q91o?list=PLLssT5z_DsK-h9vYZkQkYNWcItqhlRJLN&t=449) - What about having movies without any user giving ratings? What problem does taking mean normalization on ratings on col? (mean normalization on a user's ratings of all movies does not make much sense )\n\nBelow are the actual new videos from this new course\n\n**Binary labels: favs, likes and clicks** \n[00:00](https://youtu.be/tnIiuLQk63I?t=0) - What is the binary label problem application in recsys?\n[01:50](https://youtu.be/tnIiuLQk63I?t=110) - How to define label 1 and label 0?\n[04:04](https://youtu.be/tnIiuLQk63I?t=244) - How to move from regression to binary prediction in recsys? \n[05:16](https://youtu.be/tnIiuLQk63I?t=316) - How to modify loss function for binary labels?\n\n**Finding related items**\n[00:00](https://youtu.be/uXMa7YwDVbE?t=0) - how to find related items using features of movies or products\n03:24 - what are the limitations of collaborative filter? (cold start problem and not-using side information problem)\n\n**Collaborative filtering vs Content-based filtering** \n[00:00](https://youtu.be/UgSBaD7s5HU?t=0) - How collaboratibe filtering differ from content based filtering? (content-based filtering use side info on both users and items) \n[01:57](https://youtu.be/UgSBaD7s5HU?t=116) - What do the features or side informations of users or items look like? What's the problem to solve in order to utilize the features?\n[05:44](https://youtu.be/UgSBaD7s5HU?t=345) - How does the features of users and items can help make predictions of ratings? The key is to convert the side info of users and items into the actual features (same length) of users and items in making predictions \n[09:10](https://youtu.be/UgSBaD7s5HU?t=550) - summary\n\n**Content-based Filtering | Deep learning for content-based filtering**\n[00:00](https://youtu.be/jhgnQB7fYKM?t=0) - How deep learning (two neuralnets) help to convert side info into features (same length) of users and items?\n[03:07](https://youtu.be/jhgnQB7fYKM?t=187) - How to do the above with a single neuralnet?\n[06:02](https://youtu.be/jhgnQB7fYKM?t=362) - How to find similar items \n[08:09](https://youtu.be/jhgnQB7fYKM?t=489) - Sum up: deep learning can easily merge two neuralnets together; how to in practice make the computation of large amount of side info of users and items into actual features fast\n\n**9.10 Advanced implementation | Recommending from a large catalogue**\n[00:00](https://www.youtube.com/watch?v=nNJPU5fwc8E&list=PLxfEOJXRm7eZKJyovNH-lE3ooXTsOCvfC&index=123&loop=0) - Why it is a challenge to make predictions when there are millions of users and items?\n[01:17](https://youtu.be/nNJPU5fwc8E?list=PLxfEOJXRm7eZKJyovNH-lE3ooXTsOCvfC&t=77) - How to make use of features of users and items to recommend for existing user daily? Retrieval of related items for the user\n[04:03](https://youtu.be/nNJPU5fwc8E?list=PLxfEOJXRm7eZKJyovNH-lE3ooXTsOCvfC&t=243) - How to use the model to predict on the retrieved items in order to speed up or ease the computation? Rank by using the neuralnet model\n[05:26](https://youtu.be/nNJPU5fwc8E?list=PLxfEOJXRm7eZKJyovNH-lE3ooXTsOCvfC&t=326) - How to decide how many items to select in retrieval step?\n[07:09](https://youtu.be/nNJPU5fwc8E?list=PLxfEOJXRm7eZKJyovNH-lE3ooXTsOCvfC&t=429) - sum up\n\n**9.11 Advanced implementation | Ethical use of recommender systems**\n[00:00](https://youtu.be/GGnPWaNjaGM?list=PLxfEOJXRm7eZKJyovNH-lE3ooXTsOCvfC) - goals of recommendation system? what are the suspicous ones?\n[04:28](https://youtu.be/GGnPWaNjaGM?list=PLxfEOJXRm7eZKJyovNH-lE3ooXTsOCvfC&t=268) - Ad examples\n[07:51](https://youtu.be/GGnPWaNjaGM?list=PLxfEOJXRm7eZKJyovNH-lE3ooXTsOCvfC&t=471) - other problematic cases", "2021972": "thank you for sharing", "2022448": "You are welcome", "2022796": "It is helpful, thanks a lot.", "2026389": "Outstanding remark... Thanks a lot!", "2051806": "Thank you very much for sharing!" }, "source": "meta" }