Apprentice: Machine Learning made Accessible

Machine learning sandbox using React and Flask. Built at HackMIT 2017

Running Apprentice

1. Install MongoDB

Start the mongo server

sudo service mongod start

Create a db called apprentice

mongo
> use apprentice
> db.users.insert({})

2. Start the Flask API

cd api
export FLASK_APP=app.py
flask run &

3. Start the application

cd app
npm install
npm start

The application will run on localhost:3000

API

POST /upload

Uploads a custom dataset

  {
    "files":[(X, ...), (y, ...)],
    "predict": <id if uploading predict dataset>/null
  }

POST /example

Specifies an existing example dataset

{
  "example": "iris"
}

POST /fit

Fits the specified learner on the chosen datset

  {
    "data": {
      "id": <id>,
      "algorithm":["svm", {<params>}]
    }
  }

POST /predict

Generates a prediction using input

  {
    "id":<id of orig dataset>
  }