Student research opportunities

WWW for Machine Learning

Project Code: CECS_841

This project is available at the following levels:
Summer Scholar
Please note that this project is only for undergraduate students.

Keywords:

Machine Learning, Pattern recognition, Algorithms, RESTful web services

Supervisor:

Dr Christfried Webers

Outline:

Many implementations of Machine Learning algorithms are only available
for specific operating systems, programming languages and data formats.
This restricts users of Machine Learning and makes it harder to publish,
verify and compare algorithms in Machine Learning.

The specification of the Protocols and Structures for Inference (PSI) project provides
resource-oriented RESTful web services which aim to overcome those limitations. The
vision is to allow users of Machine Learning to access algorithms and data without
demanding detailed knowledge about implementation details.

Today, we can use the World-Wide-Web without knowledge of the programming language or the operating system the server is implemented with. Tomorrow, we should be able to apply Machine Learning to our data without having to know the intricate details of the implementation. This project will build an open-source library and provide a number of Machine Learning algorithms as a web service.

Goals of this project

This project aims to implement a number of Machine Learning algorithms used in the
course "Introduction to Machine Learning" (COMP4670/64670) and data descriptions
using the PSI services. It will lay the foundation for a growing number of algorithms
made available to future students of Machine Learning and a public audience. It will
utilise and enhance a library dealing with http communication and the encoding, decoding
and verifying of JSON data/schema.

Requirements/Prerequisites

Strong background in basic Machine Learning algorithms. Strong programming skills in C. Knowledge of RESTful web services an advantage, but can also be acquired while doing the project.