Exploratory data mining

You are asked to use OpenRefine, Weka to conduct an exploratory data mining of
this data, and to produce a SHORT report about what you discover from the data.
Tasks and mark allocations are as follows

  1. Prepare and clean the data for analysing. At this stage, you are expected to
    undertake the following procedures: Understand Data and Prepare Data. At
    the preparation stage, you should clean, and convert the data from the
    XLSX format into the ARFF format that can be accepted by WEKA. This
    includes transforming data from one type to another in order to use some
    particular algorithms. For example, you might need to transform values of
    a particular attribute from nominal to numerical in order to use a
    regression algorithm, or from numerical to nominal in order to use
    association algorithms. Therefore, you are expected to prepare more than
    one version of data for the analysing.
    [20%]
  2. Analyse the data with appropriate techniques/algorithms such as
    Classification, Regression, Association and Clustering algorithms. At this
    stage, you should find out some interesting patterns, such as which kind of
    applicants are likely to be safe to offer loans to, does skilled residents have
    any advantage? Will applicants’ credit history help? It would be fine if
    algorithms from any three of the four categories are used: Classification,
    Regression, Association and Clustering. These patterns should be
    represented by rules (about 6 rules from each of the algorithms used),
    supported by statistical information, such as accuracy and coverage, if
    possible.
    (65%)
  3. Based on your analysis in 2, summarise the overall findings in the data.
    Critically discuss and compare the algorithms used and draw an overall
    conclusion with justification about which techniques are most effective for
    making discoveries and gaining insights into the data.
    (15%)
    Total [100%]
find the cost of your paper

This question has been answered.

Get Answer