Recommendation system based on previously registered data

Question

Recommendation system based on previously registered data

Navigation

#1 by (1 votes)

0

I am a layman in machine-learning I started a course in Alura about system class using pip however does not solve my problem, I would like a library for a system recommendation , I would collect data related to each game of an electronic game and recommend the best route for the player to follow.

The logic I develop, I would just like some recommendation (A North) of libraries and / or languages to develop this.

EDIT: I will apply machine-learning in a game, at the end of each game the player would fill in data such as: where he fell on the map, how long he was alive, how many players he killed and with which weapons and especially where he was in the match. With this I would train, analyze, or whatever an algorithm to give tips to him for example where he should fall, which weapons he should use and other things.

machine-learning

asked by anonymous 11.12.2017 / 13:36

1 answer

Execute a procedure while the component is pressed PHP Mailer class does not work

score 1 · Accepted Answer

Are you sure your issue is a recommendation? It fits into recommendations problems where the number of items that can be recommended is too large and therefore can not be resolved in a scalable way by common classification algorithms. In your case, it seems to me more of a multi-class problem, where the number of possible routes is not very large.

For a recommendation, a common approach is to use algorithms based on the singular value decomposition of the interaction matrix. This package in R has a pretty cool implementation of a bunch of such algorithms. At Spark you have an implementation of ALS that seems to be scalable. In python, there are also several options. I really enjoy LightFM . But there is also Surprise .

Otherwise, most classification algorithms also have extensions for multiple classes. One-vs-Rest is usually done, an algorithm is trained to classify each class versus the others, and then the class with the highest score is chosen.