Inspiration
When we were young boys in the streets of Jihlava city, we never had any money and therefore we never bought anything. Now we've focused to create a recommendation platform, so we at least know what to buy. Young passionate programmer Martin aka The Rat has dreamed of programming machine learning, which led him to ask his friend Filip, who is quite experienced in this area (but still quite bad). They were joined by Jáchym studying his first year of computer science degree, so he could learn something new.
What it does
The code aims to recommend products to the user based on previous of him and other users' activity. Based on that it creates a similarity matrix between users suggesting them to buy what other people viewed.
How we built it
Using Python and sparse matrix data representation. We've tried to utilize google cloud and big query, but unfortunately, we are not experienced enough to make it run.
Challenges we ran into
We've started with pandas DataFrames data representation, but unfortunately, this solution was not feasible with the memory of our computers, even after hashing the hexadecimal indices. We are not that proficient in the big data approach.
Accomplishments that we're proud of
We have at least finished the algorithm...
What we learned
It is the first time we've used sparse matrices, google big query and many more technologies that we haven't used before.
What's next for our project?
Apparently, our future doesn't lie in a groundbreaking recommendation system, we might continue with going to pubs in Jihlava and watering flowers. We will fight against Mandelinka instead of RAM problems.
Log in or sign up for Devpost to join the conversation.