The number of the test-set examples that were correctly classified.
and interpreter, as you did with the previous homeworks. command-line arguments and invokes the appropriate source-code program should make a small script called bayes that accepts the a language that is not compiled to machine code (e.g. Java), then you If you are using
<train-set-file> <test-set-file> <n|t> the header section. Your program can assume that the class variable is named 'class' and it is the last variable listed in Example ARFF files are provided below. Your program needs to handle only discrete variables, and simple ARFF files (i.e. don't worry about sparse ARFF files and instance weights). See the link above for a brief, but more detailed description of the ARFF format. Lines starting with '%' are comments. Each ARFF file starts with a header section describing the variables and the class labels. class label of the instance. values are separated by commas, and the last value on each line is the this format, each instance is described on a single line. The variable Your program should read files that are in the ARFF format. In
where the last argument is a single character (either 'n' or 't') that indicates whether to use naive Bayes or TAN. command-line arguments as follows:
bayes The program should be called bayes and should accept four
site. the course Canvas file as Homework #4 at named You can use a t-test calculator, such as this one for this exercise. resulting p value. accuracy. Report both the value of the t-statistic and the to determine the statistical significance of the difference in the models achieve for each fold and then use a paired t-test partitioning of the data set for both algorithms. Report the accuracy data set to compare naive Bayes and TAN. Be sure to use the same the chess-KingRookVKingPawn.arff For this part, use stratified 10-fold cross validation on <Wisc username>_hw4.zip. Upload this zip as well as a file called hw4.pdf that shows your work for Part 2. All files should be compressed as one zip file management system. Turn in all source files and your runnable program You should turn in your work electronically using the Canvas course Submitting Your Work
Part 2
data set. This directory contains the outputs your code should produce for each You can test the correctness of your code using lymph_train.arff and lymph_test.arff, as well as vote_train.arff and vote_test.arff.