javi98/python-ml-portcanto repository overview

⁠Python Machine Learning Portcanto

Portcanto is a project to simulate a bicycle journey. 4 types of cyclists have been defined that differ in the time it takes to complete the journey.

The objective is to discover the 4 patterns with the KMeans clustering algorithm.

We want to create synthetic data in order to perform an analysis of the data with AI (basically a clustering problem).

It consists of two main files:

• generardataset.py: It consists of generating different data for each type of cyclist and storing all the data in a csv file.
• clustersciclistes.py: It consists mainly of performing an exploratory analysis of the data in the dataset, generating an image to be able to visualize the different clusters, generating reports, classifying new data, etc.

It also consists of different folders and scripts related to the analysis of the data and the AI ​​solution to the clustering problem.

Github repository: https://github.com/JAVI-CC/python-ml-portcanto⁠

⁠Setup

$ git clone https://github.com/JAVI-CC/python-ml-portcanto
$ cd python-ml-portcanto
$ docker compose up -d
$ docker compose exec app python generardataset.py
$ docker compose exec app python clustersciclistes.py

⁠docker-compose.yml

#version: '3'
services:
  app:
    build:
      context: .
      dockerfile: Dockerfile
    image: javi98/python-ml-portcanto:1.0.0
    container_name: python-ml-portcanto-app
    restart: unless-stopped
    ports:
      - 5000:5000
    #tty: true
    #stdin_open: true
    command: mlflow ui --host 0.0.0.0
    volumes:
      - ./:/code

Once you have the container deployed, You can access the MLflow web interface through the following url: http://localhost:5000⁠