in the diabetes.arff file in the weka:

1/ Prepare a box plot diagram (box and whiskers) diagram for the Plasma glucose concentration in PID (second attribute) /// I think by using excel

2/ List the number of 0s for each applicable attribute and calculate the %. Present the answer in a table.

Attribute	Number of values	Number of 0 zero values	% of 0 values
Plasma
BMI
insulin	768	374

3/ Since BMI should be > 0, what is represented by 0 in the PID data set? Check other attributes.

4/ Weka uses ? to indicate missing values. Place ? for the missing values in PID set and reopen the file in Weka. There are many methods to handle the missing values. We will use a simple imputation method: replacement of the missing value by the mean value. Use the pre-processing filter (unsupervised/attribute) ReplaceMissingValues to replace missing values with the means. Discuss the problem with imputation used for skin fold thickness and serum insulin.