Develop and Assess Unsupervised Anomaly Detection Methods

This project was completed as a intern opportunity with HPCC Systems in 2019.

Project Description

Create an ECL Bundle encompassing several widely used Anomaly Detection algorithms. Identify at least one algorithm that is parallelizable and implement efficiently in ECL:

Identify a publicly available dataset and method for assessing anomalies in the data
Research state of the art Anomaly Detection algorithms and choose two or more for implementation
Implement methods in ECL on HPCC Systems cluster, and assess results in identifying target anomalies.

By the mid term review we would expect you to have:

To have implemented atelast one algorithm and tested it.

Mentor	Roger Dev Contact Details Backup Mentor: TBD Contact Details
Skills needed	Knowledge of ECL. Training manuals and online courses are available on the HPCC Systems website. Knowledge of distributed computing techniques Knowledge of anomaly detection algorithms
Deliverables	Checked in code for the implementation in github Determine an appropriate dataset for testing the algorithms. Test code demonstrating the correctness and performance of the algorithm. Supporting documentation.
Other resources	HPCC Systems website JIRA issue for this project: TBC Learning ECL documentation and on-line training courses. Examples of existing code HPCC Systems Machine Learning documentation

HPCC

Develop and Assess Unsupervised Anomaly Detection Methods

Analytics

Mentor

Skills needed

Deliverables

Other resources

Related content