In many data-centric AI approaches, data must be prepared to be in the appropriate form and some computer-readable description of the data is necessary. Hence the first part of this assignment is data preparation.
Task:
· Obtain the broadwaystratefied.csv file from my assignments page on my www page. It is a comma-delimited file (data items separated by commas, with .csv extension). If you want to look at it, double click on it and Excel should open it.
· You need to convert this file to three files to be used by the AIAI CBR program. First, create the “case base” file.
a. Copy the .csv to the samename.cbr
b. Remove the first line of the file (open in Excel or Wordpad or Notepad)
· Second, use AIAI CBR to do most of the work in creating the .key file:
a. From AIAI CBR, choose Options
b. Select the Keys/Files/Stats tab
c. In the File Management group, to the left of Templates, click on the 3rd icon “Create a Template”
d. Find the .cbr file
e. Description of the data will be guessed at by the software; you will need to make changes (double click on a value to change):
i. Operators – make sure that they are appropriate (if numbers can be a range of matching instead of all or nothing matching, make sure that “Number Fuzzy” is specified; Decide if character strings should have partial matching – and specify String Exact or String Fuzzy)
ii. Weighting – make sure that the “ID” attribute and the attribute that is the “answer” (last attribute) have weights set to 0. Any attributes that shouldn’t be used to judge similarity should have weights set to 0. NOTE – the text string describing how long the show ran directly corresponds to the last attribute (that we will be predicting) – make sure it has a zero weight!!!!! Initially, other weights should be 1
iii. Goals – make sure that the “answer” attribute has Goals = True; all others should be False
iv. Make sure you set the ID attribute using the button provided
f. Save using the diskette icon
· Third, copy the .cbr file to create a duplicate test (.tst) file
· Turn in the files that you created (soft (electronic) copy)
· Run the AIAI CBR, after loading the case base (.cbr) file and the test (.tst) file (make sure you load the .key file if you left AIAI CBR after creating it). Use “Run Batch CBR” and save the results in a file with a descriptive name (including your name as part). This serves as a baseline for future work. Turn in this results file, and answer the following questions:
a. What do you think the “chance” probability of getting a prediction correct is in this problem? Explain.
b. Do you think that the programs’ accuracy is significantly better/worse than “chance”? Explain.
a. Do you notice a general curve to the results? What do you think is happening?
b. Do you think that the programs’ accuracy is significantly better/worse than “chance”? Explain.
c. Do you notice anything interesting about the predictions at the point where accuracy is highest?
a. What appears to work well? What do you think could be improved?
Turn In:
· Softcopy of your created files (preferably named something related to them + you)
· Hard or softcopies of your results files
· Answers to all questions above (either hard or soft copy).
NOTE: