Online Knime, Logistic regression, Data tutor - Bahawalpur, Pakistan - TeacherOn

    TeacherOn
    TeacherOn Bahawalpur, Pakistan

    Found in: beBee S2 PK - 1 week ago

    Default job background
    Part time
    Description
    URGENTLY NEED SOMEONW WHO KNOWS HWO TO DO KNIME


    This is the desc:
    For each analysis, partition the data set into training (70%) and test (30%) sets through "Stratified
    sampling". Please make sure to select "Use random seed".


    • Generate a classification (i.e., decision) tree model based on your training data set and test
    it on your test data set.

    Use the following settings for the decision tree:

    o Quality measure:
    Gini index


    o Pruning method:

    MDL
    o Check reduced error pruning
    o Min number records per node: 2


    • Generate a logistic regression model based on your training data set and test it on your test
    data set.


    • Conduct a K-nearest neighbors' analysis (k=7).
    • Compare the performance of the three approaches (i.e., classification tree, logistic
    regression, k-nearest neighbors). Which one do you recommend GCC to use and why?
    (The answer to this question should be included in your managerial report)

    What to submit:

    • One KNIME workflow for each of the models/analyses above (you should submit three
    KNIME files). To do so, you should go to File- Export KNIME Workflow. When exporting
    your models/analyses, do NOT reset workflow(s) before export (i.e., this option should be
    unchecked).

    • An MS Word document in which you include a snapshot of each of your KNIME workflows
    above as well as snapshots of table(s) and chart(s) that show the performance of each workflow.
    Please make sure that you include headers and/or section titles for different analyses


    Level:
    Beginner


    Gender Preference:
    None


    Meeting options:
    Available online - via skype etc.
    At home - Student can meet at their place.