C3_W2_RecSysNN_Assignment_Dataset

Course Q&A Machine Learning Specialization Unsupervised Learning, Recommenders, Reinforcement
1

Hey,
I was looking through the data set of movie features and user features. I found that the user id is not unique. Multiple rows with the same user id are present. Is it by design or I’m missing something? Does duplicate data help in anyway?

[user id] [rating count] [rating ave] Act ion Adve nture Anim ation Chil dren Com edy Crime Docum entary Drama Fan tasy Hor ror Mys tery Rom ance Sci -Fi Thri ller
2 22 4.0 4.0 4.2 0.0 0.0 4.0 4.1 4.0 4.0 0.0 3.0 4.0 0.0 3.9 3.9
2 22 4.0 4.0 4.2 0.0 0.0 4.0 4.1 4.0 4.0 0.0 3.0 4.0 0.0 3.9 3.9
2 22 4.0 4.0 4.2 0.0 0.0 4.0 4.1 4.0 4.0 0.0 3.0 4.0 0.0 3.9 3.9
2 22 4.0 4.0 4.2 0.0 0.0 4.0 4.1 4.0 4.0 0.0 3.0 4.0 0.0 3.9 3.9
2 22 4.0 4.0 4.2 0.0 0.0 4.0 4.1 4.0 4.0 0.0 3.0 4.0 0.0 3.9 3.9
2 22 4.0 4.0 4.2 0.0 0.0 4.0 4.1 4.0 4.0 0.0 3.0 4.0 0.0 3.9 3.9
2 22 4.0 4.0 4.2 0.0 0.0 4.0 4.1 4.0 4.0 0.0 3.0 4.0 0.0 3.9 3.9
2 22 4.0 4.0 4.2 0.0 0.0 4.0 4.1 4.0 4.0 0.0 3.0 4.0 0.0 3.9 3.9
2 22 4.0 4.0 4.2 0.0 0.0 4.0 4.1 4.0 4.0 0.0 3.0 4.0 0.0 3.9 3.9
2 22 4.0 4.0 4.2 0.0 0.0 4.0 4.1 4.0 4.0 0.0 3.0 4.0 0.0 3.9 3.9
2 22 4.0 4.0 4.2 0.0 0.0 4.0 4.1 4.0 4.0 0.0 3.0 4.0 0.0 3.9 3.9
2 22 4.0 4.0 4.2 0.0 0.0 4.0 4.1 4.0 4.0 0.0 3.0 4.0 0.0 3.9 3.9
2 22 4.0 4.0 4.2 0.0 0.0 4.0 4.1 4.0 4.0 0.0 3.0 4.0 0.0 3.9 3.9
2 22 4.0 4.0 4.2 0.0 0.0 4.0 4.1 4.0 4.0 0.0 3.0 4.0 0.0 3.9 3.9
2 22 4.0 4.0 4.2 0.0 0.0 4.0 4.1 4.0 4.0 0.0 3.0 4.0 0.0 3.9 3.9
2 22 4.0 4.0 4.2 0.0 0.0 4.0 4.1 4.0 4.0 0.0 3.0 4.0 0.0 3.9 3.9
2 22 4.0 4.0 4.2 0.0 0.0 4.0 4.1 4.0 4.0 0.0 3.0 4.0 0.0 3.9 3.9
2 22 4.0 4.0 4.2 0.0 0.0 4.0 4.1 4.0 4.0 0.0 3.0 4.0 0.0 3.9 3.9
2 22 4.0 4.0 4.2 0.0 0.0 4.0 4.1 4.0 4.0 0.0 3.0 4.0 0.0 3.9 3.9
2 22 4.0 4.0 4.2 0.0 0.0 4.0 4.1 4.0 4.0 0.0 3.0 4.0 0.0 3.9 3.9
]
here the user Id is 2 for many entries.
2

Hi @Swastik_Sambhav

It is not the training set duplicate… this is just display function for rated each movies that watched by user j it is duplicate because it would be same for all movies that user rated it … because it compute for example how this user = 2 rated action type for all movies he watched = 4.2 and so on and note in the note book it said that like this image

image
image1386×357 22.4 KB

Thanks!
Abdelrahman

3

We have 2 sets of inputs to the neural network: 1. The user features and 2. The movie features.

For the movies rated by a user, the user features will remain the same, while the movie features will be different for each movie.

For example: User 1 has rated movie 3 and 5. So the user features for user 1 will be passed as input to the User_NN twice, corresponding to movie 3 and movie 5 features being passed on as input to Item_NN

4

Oh, Yes I understood
Thnks