For this project, you will use the MovieLens 100K Dataset consisting of 100,000 movie ratings.
This can be found at:
[login to view URL]
This is a cut-down version of the full dataset that contains 20 million ratings.
The objective is to use it to provide users with recommendations of films they may enjoy.
The data is CSV format with a variety of separators used.
The three most important files are:
[login to view URL] Each line provides the details of one user.
[login to view URL] Each line provides details of one movie.
u. item Each line represents one rating of one movie by one user.
[login to view URL] The movie genres.
Movie IDs are consistent across all files. They are consistent with the full database and may, therefore, contain gaps.
Write a Python program to parse the data files and identify the highest rated film in each genre.
If more than one film in a genre scores the same rating, provide the complete list.
Extend your code using a user based collaborative filtering approach to make an individual recommendation to a user based on the set of movie ratings they have provided.
It is suggested that you use k-nearest neighbour in order to identify the closest matches, however, you are free to use other methods so long as you explain your chosen method.
Experiment with changing the number of users that are considered similar to the target user.
What if any impact does this have on the recommendations made? You should reserve some users as test subjects.
Extend your code from Task 2 to include the user details from [login to view URL] as part of the matching process.
What if any difference does this make to the results?
You will need to think carefully about how you convert the fields in [login to view URL] into suitable parameters.
You should include a brief report that:
1. The methods used to achieve the two tasks.
2. Describes the code you've written and any issues you had developing it.
3. The results of your experimentation and any conclusions you draw from it.
The report should not be any longer than two pages.
In addition to your report you should include a file(.ipynb) containing your code (with detailed comments).
I have worked on the MovieLens data set before and I also have made a user based recommendation system and an item based recommendation system. Hence, the tasks assigned will not be problematic.
11 freelancere byder i gennemsnit £149 på dette job
Hi Its Abdullah Al Mashud,expert in machine learning.I can complete your task within your desired time [login to view URL] contact me for discussing about the [login to view URL]
Hi,i'm a data scientist with a BS degree in computer science i will do your task as fast as i can and i will achieve it exactly as you want,don't worry about any thing contact me for discussion