Assignment #1 (Due 11:59pm, 14 July 2020)
Instructions: - Answer all questions.
-Interpret the questions logically, show your steps and write down your assumption(s) when necessary.
-Please submit your answer to L@PU before the due date.
-Late Submission Policy
o3-hour “grace period” is given.
o10% off for every 3-hour late
-Plagiarism Policy
oBoth giver and receiver subject to the same penalty below
oAll the students involved not only will receive 0 marks for this assessment, but also will have an additional 50% penalty applied, e.g., 5 marks for a 10-mark assessment.
1.Suppose you are asked by the Octopus Card Company to analyze the following Card Database.
Octopus Card Database for Training
Octopus ID Day of the Week From To Transportation Taken
1 Mon-Thurs HK HK MTR
2 Mon-Thurs Kln Kln Bus
3 Mon-Thurs NT Kln Bus
4 Fri-Sun HK HK MTR
5 Mon-Thurs Kln Kln MTR
6 Mon-Thurs NT NT Bus
7 Mon-Thurs HK HK Bus
8 Fri-Sun Kln Kln MTR
9 Mon-Thurs Kln Kln Bus
10 Mon-Thurs HK HK Bus
11 Fri-Sun HK HK MTR
12 Mon-Thurs Kln Kln Bus
13 Mon-Thurs HK HK Bus
14 Mon-Thurs Kln Kln MTR
15 Fri-Sun HK NT MTR
16 Mon-Thurs NT Kln Bus
17 Mon-Thurs HK NT Bus
18 Mon-Thurs Kln HK MTR
19 Fri-Sun HK NT MTR
20 Mon-Thurs HK HK MTR
Octopus Card Database for Testing
Octopus ID Day of the Week From To Transportation Taken
21 Mon-Thurs Kln Kln Bus
22 Mon-Thurs HK HK MTR
23 Mon-Thurs Kln Kln Bus
24 Mon-Thurs Kln Kln Bus
25 Fri-Sun Kln HK MTR
Your project team member tried to construct a decision tree for the first 20 records above, i.e., the training dataset, and obtained the result as follows.
a)Write down the classification rules that can be extracted from the constructed tree.
b)Another project team member commented that the decision tree above is too deep. Prune the tree as much as possible so that the classification rate (accuracy) of the pruned tree for the testing database can be maintained, i.e., as high as that obtained in part (a).
c)If some more database records are available and added to the training database above, describe the situations for
i)reconstruction of the decision tree is NOT necessary
ii)partial reconstruction of the decision tree is necessary
iii)full reconstruction of the decision tree is necessary
2.In our lecture notes, we have already seen the derivative of a sigmoid function wrt its weight parameters as follows.
we have
Let us now consider the softmax function with the function form
.
Derive the derivative of softmax wrt . You MUST use the symbols above, i.e., , , etc., to present your answer.
Hint:
Using the quotient rule and let and , we have