Download solutions
Birla
Institute of Technology & Science, Pilani
Work-Integrated
Learning Programmes Division
Second
Semester 2016-2017
Mid-Semester
Test (EC-2 Regular)
Course No. : IS ZC415
Course Title : DATA MINING
Nature of Exam : Closed Book
Weightage : 30%
Duration : 2 Hours
Date
of Exam : 25/02/2017 (AN)
No.
of pages: 2
No.
of questions: 5
Note:
1. Please
follow all the Instructions to Candidates given on the cover page of the
answer book.
2. All
parts of a question should be answered consecutively. Each answer should start
from a fresh page.
3. Assumptions
made if any, should be stated clearly at the beginning of your answer.
Q.1 (a)
What is mode of the following data?
10, 2, 30, 14, 50 [1]
Q.1 (b)
Eleven
students were asked to measure their pulses for 30 seconds and multiply by two
to get their one minute pulse rates. The results were: 62, 32, 60, 66, 70, 72,
74,
74,
78, 80, 84. Create five-number summary for the pulse rates and draw boxplot.
[3]
Q.1 (c)
Students admitted for a certain
course have mean score of 560 and a standard deviation of 60. Calculate the z-score
of a student having a score of 500. [1]
Q.1 (d)
Calculate the cosine similarity between the two phrases below.
Feature vector of a word occurring multiple times is greater than 1. Clearly
show steps of your calculations.
mid
term regular exam
regular exam mid term mid
term regular exam [2]
Q.2. You
are given 10 training samples. They are divided into four classes: a, b, c, and d.
One sample
belongs to A, two belong to B, three belong to C, and four to D. Use the
following log2
table to answer the questions:
|
p
|
log2(p)
|
|
0.1
|
-3.32
|
|
0.2
|
-2.32
|
|
0.3
|
-1.74
|
|
0.4
|
-1.32
|
|
0.5
|
-1.0
|
|
0.6
|
-0.74
|
(a)
What
is the total information contained in the samples? [2]
(b)
What
is the total Gini index? [2]
Q.3 (a)
Given below is a database of flight delays over a period and
under various conditions. We Want to
create a decision tree classifier with information gain(entropy) as the
attribute splitting criterion.
|
Feature
|
Value = Yes
|
Value = no
|
|
Rain
Fog
Summer
Winter
Day
Night
|
Delayed=30, not Delayed=10
Delayed=25, not Delayed=15
Delayed=5, not
Delayed=35
Delayed=20, not Delayed=10
Delayed=20, not Delayed=20
Delayed=15, not Delayed=10
|
Delayed=10, not Delayed=30
Delayed=15, not Delayed=25
Delayed=35, not Delayed=5
Delayed=20, not Delayed=30
Delayed=20, not Delayed=20
Delayed=25, not Delayed=30
|
Which feature should be at the root of decision tree? [2]
Q.3 (b)
Given the following training documents and their classes:
|
Document#
|
Content of document
|
Class
|
|
1
|
good
|
Ham
|
|
2
|
very good
|
Ham
|
|
3
|
bad
|
Spam
|
|
4
|
very bad
|
Spam
|
|
5
|
very bad very bad
|
Spam
|
Use Naïve Bayes classifier with Laplace (+1) smoothing to find the class
of a document with the following contents:
very good bad very very bad [5]
Q.4. Suppose you have the following candidate
itemsets of length 4:
{1 2 3 5}, {1 2
4 7}, {1 2 5 6}, {1 3 5 9}, {1 4 5 7}, {1 5 6 9}, {2 3 5 9}, {3 4 5 9},
{4 5 6 8}, {5 6
7 9}
(a)
Use
hash function k mod 5 to create a hash tree of the itemsets. Assume that each
leaf
node can store a
maximum of three itemsets. [4]
(b)
Given
transaction {1, 2, 3, 5, 7, 9}, which leaf nodes of the hash tree will be
visited
for support-counting?
Clearly show the visited leaf nodes in the hash tree. [2]
Q.5. Given that min
support is 2, and min confidence is 70%, find all association rules from
the following market
basket dataset using Apriori: [6]
|
Transaction ID
|
Items
|
|
1
|
a, b, c
|
|
2
|
b, c, d, e
|
|
3
|
c, d
|
|
4
|
a, b, d
|
|
5
|
a, b, c
|
*********
ANSWERS
ReplyDelete