1.Choose the area of your preference, whatever you would like to describe in a dataset and explain using data mining.For example: actresses/actors, food, movies, sports, music bands, or anything you want.
Create a data file in .arff format containing about 20 entries, each described by
about 4 attributes, with the last attribute containing your preference (class attribute), e.g.
@relation food
@attribute calories numeric
@attribute taste {sweet, sour, bitter, salty}
@attribute course {appetizer, main, dessert, drink}
@attribute vegetarian {yes, no}
@attribute like_it {yes, no}
@data
100, sweet, dessert, yes, yes%icecream
80, bitter, drink, yes, yes%beer
2, sweet, dessert,yes, no%cake
Compare 3 algorithms for classification of your data: decision trees, a classification or an association rule learner, and naive Bayes. For each algorithm check what the error is (which algorithm can explain your personal liking the best), and observe the generated rules (do they tell you anything interesting?).
2.Use the following learning schemes to compare the training set and 10-fold stratified cross-validation scores of the labor data (in labor_neg_nominal.arff):
•k-nearest neighbours (IBk) with decision trees (j48.J48)
•k-nearest neighbours (IBk) with decision trees j48.J48 with option -M 3, so that each leaf has at least 3 instances.
A)What does the training set evaluation score tell you? B)What does the cross-validation score evaluate?
C)Which one of these models would you say is the best?Why?
3.Use the following learning schemes to analyze the Titanic data (in titanic.arff).
C4.5- weka.classifiers.j48.J48
Association rules-weka.associations.apriori
Decision List- weka. Classifiers.PART
A)What is the most important descriptor (attribute) in titanic.arff?
B)How well were these methods able to learn the patterns in the dataset? Quantify your
answer?
C)Compare the training set and 10-fold cross-validations scores of the methods.
D)Would you trust these models?Did they really learn what was important to survive the
Titanic disaster?
E)Which one would you trust more, even if just very slightly? Why?
4.Choose one of the following three files: soybean.arff, autoprice.arff, hungarian, zoo.arff or zoo2_x.arff and use any two schemas of your choice to build and compare the models. Which one of the models would you keep?Why?
5.Use the Association rule learner APRIORI method to find the association rule in the Weather.nominal data set.How many rules did it produce?How large are the item sets? What was the largest one? What happened when you increased/decreased the confidence level? What about the number of rules?What happens when you increase the confidence parameter to 2?Why?