Assignment Task

Objective 

The aim of this assignment is to generate value and insight from the processing of heterogeneous data. This will be achieved by implementing several analytic methods techniques algorithms, evaluating them and comparing the effectiveness of the adopted approaches. The development and implementation of the data analysis project will be supported by team-based effort and weekly meetings.

Description 

The final report should be an original and individual submission, but it will be underpinned by a group effort and by effective sharing of data and partial results. It is important that individual contributions shared among the group are clearly defined (see “Authorship Contribution” below for further details). These contributions should be agreed upfront in a designated meeting. The management of the data during (and after the project) can be described in a formal Data Management Plan (DMP). A template is available on Blackboard Learn. The DMP should be discussed with the group in a designated meeting.

The following parts should be developed by shared group effort:

• Data collection
• Data preparation and cleaning
• Exploratory data analysis

Each member of the group is expected to implement and apply at least two methods/approaches including:

• One machine learning method for prediction (regression or classification)
• One High Performance Computational technique for distributed data analysis. It is expected that an alternative method/approach to MapReduce will be used.

For both of these two methods/approaches, each member of the group should produce their own results and distribute/exchange their results among all the other members. Finally, each member is expected to independently compare, discuss and evaluate these shared results. Submissions will be graded on technical ability, creativity, practicality, and their use of concepts introduced in different study blocks, in particular CS5706 Machine Learning and CS5710 High Performance Computational Infrastructures.