Business Decision Making Assignment | Homework For You

Task:

1.    If you have Supervised data set do the followings:
a.    Measure class distribution of your data set.
b.    Select 20 instances from the data set with same class distribution.( i.g. if there are 1000 instance in your original data sets with 700 belongs to one class and remaining 300 belongs to another class than in your selected 20 instances 14 belongs to one class and remaining 6 belongs to other class).
Or
For unsupervised data set, select 20 instances and randomly assign class label to all 20 instances by following 7::3 ratio in class distribution.
2.    Construct two decision tree using info-gain and error-rate with 6::2::2 ratio for  training::validation::testing.
a.    Propose the best decision tree using MDL and Validation set with 1 bit for internal node and 2 bit for miss-classified record.
b.     Select the best decision tree using cost matrix and testing set:

a    b
a    2    -3
b    -2    3

3.    Apply 2*2 cross fold validation method and Ripper algorithm to generate rule-base and evaluate the performance of the model using Accuracy and recall of minority class.

4.    Apply 2*2 cross fold validation method and NB algorithm to generate classification model and evaluate the performance of the model using precision and f-measure of majority class. Get Statistics homework help today