/machine-learning

Web-interface + rest API for classification and regression (https://jeff1evesque.github.io/machine-learning.docs)

Primary LanguageJavaScriptOtherNOASSERTION

Machine Learning Build Status Coverage Status

This project provides a web-interface, as well as a programmatic-api for various machine learning algorithms.

Supported algorithms:

Contributing

Please adhere to contributing.md, when contributing code. Pull requests that deviate from the contributing.md, could be labelled as invalid, and closed (without merging to master). These best practices will ensure integrity, when revisions of code, or issues need to be reviewed.

Note: support, and philantropy can be inquired, to further assist with development.

Configuration

Fork this project, using of the following methods:

  • simple clone: clone the remote master branch.
  • commit hash: clone the remote master branch, then checkout a specific commit hash.
  • release tag: clone the remote branch, associated with the desired release tag.

Installation

To proceed with the installation for this project, users will need to decide whether to use the rancher ecosystem, or use docker-compose. The former will likely be less reliable, since the corresponding install script, may not work nicely across different operating systems. Additionally, this project will assume rancher as the primary method to deploy, and run the application. So, when using the docker-compose alternate, keep track what the corresponding endpoints should be.

If users choose rancher, both docker and rancher must be installed. Installing docker must be done manually, to fulfill a set of dependencies. Once completed, rancher can be installed, and automatically configured, by simply executing a provided bash script, from the docker quickstart terminal:

cd /path/to/machine-learning
./install-rancher

Note: the installation, and the configuration of rancher, has been outlined if more explicit instructions are needed.

If users choose to forgo rancher, and use the docker-compose, then simply install docker, as well as docker-compose. This will allow the application to be deployed from any terminal console:

cd /path/to/machine-learning
docker-compose up

Note: the installation, and the configuration of docker-compose, has been outlined if more explicit instructions are needed.

Execution

Both the web-interface, and the programmatic-api, have corresponding unit tests which can be reviewed, and implemented. It is important to remember, the installation of this application will dictate the endpoint. More specifically, if the application was installed via rancher, then the endpoint will take the form of https://192.168.99.101:XXXX. However, if the docker-compose up alternate was used, then the endpoint will likely change to https://localhost:XXXX, or https://127.0.0.1:XXXX.

Web Interface

The web-interface, can be accessed within the browser on https://192.168.99.101:8080:

web-interface

The following sessions are available:

  • data_new: store the provided dataset(s), within the implemented sql database.
  • data_append: append additional dataset(s), to an existing representation (from an earlier data_new session), within the implemented sql database.
  • model_generate: using previous stored dataset(s) (from an earlier
  • data_new, or data_append session), generate a corresponding model into
  • model_predict: using a previous stored model (from an earlier model_predict session), from the implemented nosql datastore, along with user supplied values, generate a corresponding prediction.

When using the web-interface, it is important to ensure the csv, xml, or json file(s), representing the corresponding dataset(s), are properly formatted. Dataset(s) poorly formatted will fail to create respective json dataset representation(s). Subsequently, the dataset(s) will not succeed being stored into corresponding database tables. This will prevent any models, and subsequent predictions from being made.

The following dataset(s), show acceptable syntax:

Note: each dependent variable value (for JSON datasets), is an array (square brackets), since each dependent variable may have multiple observations.

Programmatic Interface

The programmatic-interface, or set of API, allow users to implement the following sessions:

  • data_new: store the provided dataset(s), within the implemented sql database.
  • data_append: append additional dataset(s), to an existing representation (from an earlier data_new session), within the implemented sql database.
  • model_generate: using previous stored dataset(s) (from an earlier
  • data_new, or data_append session), generate a corresponding model into
  • model_predict: using a previous stored model (from an earlier model_predict session), from the implemented nosql datastore, along with user supplied values, generate a corresponding prediction.

A post request, can be implemented in python, as follows:

import requests

endpoint = 'https://192.168.99.101:9090/load-data'
headers = {
    'Authorization': 'Bearer ' + token,
    'Content-Type': 'application/json'
}

requests.post(endpoint, headers=headers, data=json_string_here)

Note: more information, regarding how to obtain a valid token, can be further reviewed, in the /login documentation.

Note: various data attributes can be nested in above POST request.

It is important to remember that the docker-compose.development.yml, has defined two port forwards, each assigned to its corresponding reverse proxy. This allows port 8080 on the host, to map into the webserver-web container. A similar case for the programmatic-api, uses port 9090 on the host.