Skip to content
/ rxys Public

The future of technology implemented today. Various implementations of multi-context, context-aware system of recommendation optimized using particle swarm and enhanced by ensemble of deep learning techniques.

Notifications You must be signed in to change notification settings

nkrth/rxys

Repository files navigation

rxys

Tired of stochastic gradient descent in SVD-like algorithms, we turn to more interesting ideas like Hybrid systems, Alternating Least Squares (ALS) and Implicit Feedback. Perhaps, ARIMA and LSTM ensembled with Fast R-CNN in a large stacknet. In this project, however, we use a concept from Swarm Intelligence known as PSO to optimize suggestions given to us by what once was a multi-context text generator.

Please read the docs for more on the available implementations. (docs will be made available shortly)

Note: Sorry, I've taken out the implementation files from the push queue. Trying to make them as easy to access as possible.

Usage

You can generate your own instance of an optimized recommender using:

!pip install rxys
import rxys
rxys.init()
rxys.sample()

This works on Python 3.7.x and backward compatibility is not supported. If this doesn't work, our PyPI is broken. It will be up shortly.

The current instance of the package is available at: https://pypi.org/project/rxys/

Related

Easier K-Means Movie Recommender: https://github.com/xueharry/CS51-Final-Project/tree/master/code This project is related to https://github.com/NicolasHug/Surprise and has implementations like Amazon Product Recommendations, Facebook Friend Suggestions and Book Suggestions as in https://github.com/dorukkilitcioglu/books2rec

Contact

For any suggestions, feedback or clarifications, reach me through e-mail at nick.kartha@gmail.com (cc: kartha@vivaldi.net for faster response)

About

The future of technology implemented today. Various implementations of multi-context, context-aware system of recommendation optimized using particle swarm and enhanced by ensemble of deep learning techniques.

Topics

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published