Collaborative Filtering (Memory Based)|Item and User based collaborative filtering recommendation

Collaborative Filtering(Memory Based)|Item and User based collaborative filtering recommendation
#CollaborativeFiltering #UserbasedCollaborativeFiltering, #UnfoldDataScience
Hello,
This is Aman and I am a Data Scientist.
About this video:
In this video, I explain in detail about collaborative filtering memory based approaches like item based collaborative filtering and user based collaborative filtering.
I explain all the fundamentals of collaborative filtering in this video and explain how collaborative filtering works.
Below questions are answered in this video:
1. What is memory based collaborative filtering?
2. What is item based collaborative filtering?
3. What is user based collaborative filtering?
4. How is similarity calculated in collaborative filtering?
5. How does collaborative filtering work?
About Unfold Data science: This channel is to help people understand basics of data science through simple examples in easy way. Anybody without having prior knowledge of computer programming or statistics or machine learning and artificial intelligence can get an understanding of data science at high level through this channel. The videos uploaded will not be very technical in nature and hence it can be easily grasped by viewers from different background as well.
Join Facebook group :
groups/41022...
Follow on medium : / amanrai77
Follow on quora: www.quora.com/profile/Aman-Ku...
Follow on twitter : @unfoldds
Get connected on LinkedIn : / aman-kumar-b4881440
Follow on Instagram : unfolddatascience
Watch Introduction to Data Science full playlist here : • Data Science In 15 Min...
Watch python for data science playlist here:
• Python Basics For Data...
Watch statistics and mathematics playlist here :
• Measures of Central Te...
Watch End to End Implementation of a simple machine learning model in Python here:
• How Does Machine Learn...
Learn Ensemble Model, Bagging and Boosting here:
• Introduction to Ensemb...
Access all my codes here:
drive.google.com/drive/folder...
Have question for me? Ask me here : docs.google.com/forms/d/1ccgl...

[@brown_bread]

at 4:50, Implicit rating definition is described incorrectly. For implicit ratings, users are not asked to give any ratings-we just observe their behavior.

[@UnfoldDataScience]

HI Himanshu, Thanks for the feedback. Did I confuse between Implicit and Explicit - YEs. Let me pin your comment on top of video for everyone's benefit. Thanks for pointing out. You guys are hero :)

[@brown_bread]

@@UnfoldDataScience Keep up the good work.

[@homlaxmigurung168]

OMG! Why you teaching us wrong? 😢😭😢

[@Birdsneverfly]

@@homlaxmigurung168 Every human and algorithm will make mistakes.

[@UnfoldDataScience]

Access Hindi, English courses here- www.unfolddatascience.com/s/store Plz register on the website

[@AidanScudder]

thanks dude, appreciate the well explained video

[@UnfoldDataScience]

My pleasure :)

[@sangeeth8086]

The way of your explanation is brilliant. Keep going 👏 👏

[@UnfoldDataScience]

Thank you Sangeeth for motivating through comment.

[@kamran_desu]

Very nice explanation. Can you clarify more on Item based. At @14:30 you mentioned that we first find similar Items rated by the User and then you mentioned ratings for other 'horror' movies. Now my confusion here is that you're talking about an attribute of the item i.e. genre and that makes this content-based not collaborative. Is the Item based simply = find similar Items to what the User has rated well, rank them based on similarity, and recommend whatever they have not watched?

[@anastasiiasnk7324]

Hello, Aman. Thanks for your video? I don't understand is it possible to use terms like user and item based for model based approach?

[@preranatiwary7690]

Awesome video again :)

[@UnfoldDataScience]

Thank you. Keep watching :)

[@azamtariq2559]

my most Favret youtuber

[@NikhilWhiskyKumar]

nc explanation...keep it up

[@UnfoldDataScience]

Thanks Nikhil. Happy Learning. Tc

[@sadhnarai8757]

Very nice Aman

[@UnfoldDataScience]

Thank you

[@evazhou2925]

Hi Aman - great video! I have a question - In the memory based CF, how do we deal with the missing ratings? For example, if the user A found the same ratings on movie i from user B and user C, and we want to recommend the move ii to user A. But user B also has null value for move ii. And user c has rating 4 for move ii. Are we going to remove the row User B? Similarly when we run the Correlation in Python, will it automatically deal with the those missing rating? Thanks a lot!

[@UnfoldDataScience]

That is a good question Eva, when we prepare data for model training, we impute the missing ratings. So, for NULL values we impute the rating. Next comes, how to impute? One way is, If the movie falls in "Horror" Genre then we take average of all the "Horror" genre movie rated by user and put in place of NULL. A generic approach. There can be more specific approaches as well.

[@shriniwasbharadwaz1183]

Very helpful video

[@UnfoldDataScience]

Glad to hear that Shri.

[@CanvasLid]

this was very helpful thank you so much

[@UnfoldDataScience]

Glad it was helpful Noor :)

[@sonal008]

if no.of users> no. Of item then we build item based recommendation engine and vice-versa for user based?

[@rajuthapa9005]

great

[@UnfoldDataScience]

Thanks Raju. Happy Learning!

[@Djrl12]

Great Video! I have a question. Is it possible to do a combination of item and user based collaborative filtering at the same time? If yes, in simple words, how would this be implemented?

[@UnfoldDataScience]

Yes an hybrid approach is possible. I can explain in detail in a separate video. Thanks for the ask. :)

[@mohamedaminenebri9372]

Hi, Your courses amazing, but please may do sub titles in English, it's helpful Thank you

[@mosama22]

Hi Aman, thank you so much for one more beautiful video, but quick question please: so can we say for simplicity that the CF approach is a classification approach where we group similar users / items together, and based on that we start to guess the user's (whatever we wanna recommend) based on what class he belongs to?!

[@UnfoldDataScience]

Hi Osama, Loosely yes, but if you see in machine learning there are many jargons already here and there hence better to be specific. For understanding yes, for explaining to someone use collaborative filtering only.

[@amritachoudhury400]

@@UnfoldDataScience what Osama mentioned we call that Clustering right .Isnt CF different ?

[@akshatachaudhari3775]

thank you for the video sir. Got a deep understanding. i just have one doubt. What's the difference between content based filtering and item based collaborative filtering??

[@UnfoldDataScience]

It will be a long answer, please got through this once: stackoverflow.com/questions/16372191/whats-difference-between-item-based-and-content-based-collaborative-filtering

[@saicharan4016]

Short, Clear and Crisp info about every topic keep more videos and playlist list them if possible ....very nice content , thanks a lot for making such videos

[@UnfoldDataScience]

Thanks Sai, your comment mean a lot to me.

[@michaeledison1974]

I need help with collaborative filtering, can I contact you to display the problem, please?

[@vinayakg9180]

good work keep doing. I was digging for a solution related to recommending the courses, which can be used by any online training platform. came across your video. thumps up, who all reading this. please do more experiments as you can, after learning doesn't just keep learning.

[@UnfoldDataScience]

Great suggestion Vinayak.

[@sarabhian2270]

I have dataset in which user_id , value of product and catagory of products is given , and I have to recommend top 3 products to user based on his older purchase ... how I can approach this problem ?

[@UnfoldDataScience]

Answered.

[@shriniwasbharadwaz1183]

Sir, you said you will implement this in python. I can't find it in your playlist

[@UnfoldDataScience]

Will upload soon Shriniwas. Occupied with few plans now.

[@sandipansarkar9211]

great explanation. Need to practice on code other wise no use

[@UnfoldDataScience]

Yes Sandipan, otherwise we forget :)

[@chodeyalekhya8204]

Sir I want code for movie recommedations system using content based filtering.

[@UnfoldDataScience]

Link in Description or in "About" section of my channel.

[@kamran_desu]

Very good explanation, is there also a separate video for detailed content based and hybrid approach?

[@UnfoldDataScience]

Thank you. Please look in this playlist. kzbin.info/www/bejne/oKbUoYeXnsp8oaM

[@hariharibolll3459]

jai shree RAM

[@coolmantej50]

It is explicit rating and not implicit rating at 5:20

[@UnfoldDataScience]

thanks for pointing out Mantej. This explicit/implicit confuses me sometimes. Thanks a lot. :)

[@kmanojus]

Can you pls. share code base link or github link?

[@UnfoldDataScience]

drive.google.com/drive/folders/1XdPbyAc9iWml0fPPNX91Yq3BRwkZAG2M

[@satyapragyandas811]

let's say i have time instead of ranking: 1.should i normalise the user-item matrix 2.how to evaluate such a memory based collaborative filtering

[@UnfoldDataScience]

How time will fit in this approach? I want to understand time of what event?

[@codeloki]

Time will not give you anything, check the frequency of an item bought.. check if you can get price.

[@satyapragyandas811]

@@UnfoldDataScience time consumed by the item. Please the 2nd question first, i want to evaluate my recommender system. how should i do it?

[@satyapragyandas811]

@@codeloki can you please tell how to evaluate such model?

[@satyapragyandas811]

@@UnfoldDataScience can you tell how to evaluate such model?

[@homlaxmigurung168]

You're so handsome I like you so much.

[@UnfoldDataScience]

Thanks! 😃

[@homlaxmigurung168]

@@UnfoldDataScience Please let me know how I could be in contact with you?