recsys_training

Recommender System Training Package

Description

Hands-on Training for Recommender Systems developed for Machine Learning Essentials 2020.

Installation

In order to set up the necessary environment:

create an environment recsys_training with the help of conda,
```
conda env create -f environment.yaml
```
activate the new environment with
```
conda activate recsys_training
```
install recsys_training with:
```
python setup.py install # or develop
```

Docker

Make sure you have docker and docker-compose installed.

Build the image with using the Dockerfile in docker

docker build -t recsys-training:mle -f Dockerfile .

Start the container with docker-compose pointing to the yaml-file
```
docker-compose up -f docker/docker-compose.yaml
```

The jupyter lab port 8888 will be mapped to the same port on your host machine, simply got to your preferred browser and enter via http://localhost:8888/

Usage

There are 9 notebooks within notebooks/ each starting with a number followed by _e_ for exercise. Within notebooks/solutions/you will find all notebooks with a solution proposal implemented. It is strongly advised to go through the notebooks in numerically ascending order.

We use MovieLens 100k as example dataset for the lessons. You can find the data in data/raw/.

Note