Reference no: EM132479389

CS5608 - Big Data Analytics - Brunel University

Assessment - Design, implementation and evaluation of a complete data analytics solution

OBJECTIVE OF THE ASSESSMENT
The aim of this assessment is to design, implement and evaluate a complete data analytics solution with the R software environment for statistical computing and graphics. You will have the opportunity to use the analytical features offered by R (through built-in functions and CRAN packages) to perform basic and advanced analytical routines, to explore the machine learning methods presented during the lectures and to use these methods for data transformation, integration, analysis and knowledge discovery. You are expected to perform all the data analysis in R and to include the scripts you have used in your report.

DESCRIPTION OF THE ASSESSMENT
You will have to use multiple data files (at least two different data sets) from existing open data sources (examples of data sources were provided in the lecture during Week 17). The task is to integrate these data files, perform a joint analysis on them and create a narrative about the data you have analysed driven by a research question of your choice. The research question should be relevant to the domain of knowledge of the data. You will have to select the most appropriate data analysis strategy to answer the research question, but you should include at least two of the methods presented in the lectures. The use of methods not presented during the module will not contribute to the grade, also if the method is relevant for the analysis.

You are invited to report the use of at least two machine learning methods (either for different tasks required to answer the research question or in a comparative fashion). The use of methods with different goals (e.g. regression, classification and clustering) is encouraged. You should include clear justification for each choice of method and data analysis technique. Data analysis results should be presented with tables and graphs.

Learning Outcome 1: Implement appropriate analytic methods/techniques/algorithms for generating value and insight from the (real-time) processing of heterogeneous data.

Learning Outcome 2: Critically reflect on analytic methods/techniques/algorithms, their ability to deliver accurate predictions of various kinds and the value and limitations of prediction.

The coursework will be marked according to the following criteria:
i) Identifying a data analytics problem and formulating a relevant research question and plan
ii) Preparing and integrating multiple data sets that are suitable to answer the research question
iii) Implementing and executing a complete and coherent data analysis in R
iv) Critically reflecting on the results of the data analysis (accuracy, limitations and interpretation)

Attachment:- Big Data Analytics.rar