• Solve Exercise 8.3 of your textbook (page 247). Include your solution in your written report (and not in your oral report).

*** You must use the Project 6 Template provided for your written report. *** (if you prefer not to use Word, you can copy and paste this format in a different editor as long as you respect the stated page structure and page limit.)

The font size should be no smaller than 11pts. Do not exceed the page limit.

• Machine Learning Technique(s): Use the following Instance-based Learning and Regression methods implemented in the Weka system and in R, or implement your own code.
  For IB1, IBk, and LWL, experiment with both nominal and numeric targets.

  • IB1: nearest neighbor classification
  • IBk: k-nearest neighbors classification. Experiment with several values of k.
  • Linear Regression
  • LWL: Locally Weighted Learning. Experiment with different classifier options (within LWL). In particular experiment with Locally Weighted Linear Regression. [In order to run locally weighted linear regression using Weka, use LWL (locally weighted learning) from the Weka's lazy classifiers, and select "Linear Regression" for the LWL's classifier option]

• Dataset(s): In this project, we will use one dataset:

  • The U.S. News & World Report Best Colleges Rankings 2010 available from The Ohio State University - Institutional Research and Planning.

    Note that:

    • The first sheet/tab ("Sheet1") of the excel spreadsheet constains a description of the contents of each sheet/tab in the file. I suggest we use the data in the "2010" sheet/tab for the project.
    • Add a new column in the dataset with the state code (e.g., "MA" for "Massachusetts"). The state code is available next to the name of the school.
    • Ideally, add any other interesting and/or relevant information about the universities not present in the dataset that you can collect from free, reputable online sources. Cite your sources.

  • Alternatively, you can collect current (2014) data by putting together data from the following online sources:

    • The The U.S.News National Universities Rankings (2014)
      From this online source collect the national universities ranked from #1 to #201 by the U.S.News. Use the "Table View (More Data)" to collect, for each university, school's name, tuition and fees, total enrollment, Fall 2013 acceptance rate, average freshman retention rate, and 6-year graduation rate. Include also a nominal attribute stating whether the university is public or private; and another nominal attribute stating which U.S. state the university is located in.

    • The "How to Understand SAT Scores in College Admissions Data An Explanation of 25th / 75th Percentile SAT Scores Found in College Profiles" By Allen Grove.
      From this online source collect the 25 and 75 percentiles (i.e., first quartile and 3rd quartile) SAT Reading, Math, and Writing Scores for as many national universities that you collected above from the U.S.News.
      You may use another free, reputable online source to collect this SAT data if you wish.

    • Optional: Any other interesting and/or relevant information about the universities (e.g., faculty-student ratio, male-female ratio, ...) that you can collect from free, reputable online sources. Cite your sources.

    [Note: The above data collection description was inspired by the University Data Set available at the UCI Machine Learning Repository. However this dataset is dated (1987) and therefore you'll be collecting similar but more recent data for this project.]

• Additional Information:

  • Target Attribute: Run a set of classification experiments using a nominal target attribute of your choice, and then run a separate set of regression experiments using a numeric target attribute of your choice.

  • Training and Testing Instances: You may restrict your test set to about 100 data instances (not used for training) to reduce the time taken by the experiments using these lazy methods.