Data Mining and Machine Learning
CSE · Final · Spring 23
Section All
Read the questions as text
Read off the PDF by AI, so it may have mistakes. The PDF is the original.
Daffodil International University
Department of Computer Science and Engineering
Faculty of Science and Information Technology
Final Examination
Semester: Spring 2023
Section: All
Course Code: CSE321
Course Title: Data Mining and Machine Learning
Time: 1.5 hours
Full Marks: 40
[N.B. Answer all questions. Marks of all questions have been indicated on the right side of the question paper.]
1. An artificial neural network has been trained on a binary classification dataset, whose target variables are red and green, encoded as 0 and 1, respectively, prior to training the network. In this network, the hidden nodes are activated using Rectified Linear Units (ReLU), and the last node is activated using Sigmoid. The trained network is shown below. Classify the data shown in Table 1 using this network. (CO4, 10)
[Figure: A neural network diagram with input units F1, F2, F3, bias, hidden units H1, H2, H3, and output unit O1 with weights and biases.]
Table 1: Test data
F1 | F2 | F3
2.1 | 1.9 | 1.02
3.4 | .04 | 1.2
2. Below is a dataset with three features: Age, Annual Income, and Spending Score. Using a clustering approach, identify the number of clusters, boundary points, and outliers in the dataset. (CO2, 5+5+5)
Age | Annual Income | Spending Score
19 | 15 | 39
21 | 15 | 81
23 | 16 | 77
31 | 17 | 40
21 | 600 | 1
3. Consider the transaction database shown in the following table. (CO3, 15)
ID | Items
1 | A, B, C, D
2 | G, B, C, D
3 | A, C, D, G
4 | A, B, F, H
5 | B, D, N, E
Let minimum support be 50% and minimum confidence be 65%. Predict all frequent item sets and the association rules.