UNIT-I: Data Warehouse & OLAP
Submission Date 12th April
Q1. Define a Data Warehouse. How does it differ from a traditional Operational Database
(OLTP) . Explain the four main characteristics of a Data Warehouse.
Q2. Sketch and explain the typical Three-Tier Data Warehouse Architecture.
Q3. Design Task: A retail store wants to track sales. Design a Star Schema involving a
Fact table (Sales) and four Dimension tables (Product, Time, Branch, Location).
Q4. Define OLAP. List and explain the five basic OLAP operations (Roll-up, Drill-
down, Slice, Dice, and Pivot) with diagrams. Explain the architectural differences
between ROLAP, MOLAP, and HOLAP.
Q5. Numerical/Logic: Given a 3D data cube with dimensions: Product (5 items), Time
(4 quarters), and Location (10 cities). How many cells are in the cube? If you perform a
"Slice" on one specific city, what are the dimensions of the resulting sub-cube?
UNIT-II: Data Mining Fundamentals & Preprocessing
Submission Date 12th April
Q1. Define Data Mining. Discuss the "Knowledge Discovery in Databases" (KDD)
process steps. Explain the different Data Mining Functionalities, specifically
distinguishing between Descriptive and Predictive mining.
Q2. Describe the integration of a Data Mining system with a Data Warehouse. What are
the benefits of this coupling? List and explain three major challenges/issues currently
facing the field of Data Mining.
Q3. Why is Data Preprocessing required? What are the consequences of mining "dirty"
data? Q6. Explain Data Cleaning techniques for handling missing values and noisy data
(Binning, Regression, Outlier analysis).
Q4. Describe Data Transformation methods, focusing specifically on Min-Max
Normalization and Z-score Normalization. Given the data set [12, 24, 36, 48, 60],
normalize the value 36 using Min-Max normalization to a range of [0, 1].
Q5. What is Data Reduction? Explain the concept of Numerosity Reduction and
Dimensionality Reduction. Explain Concept Hierarchy Generation for categorical data.
Provide an example using "Geography" (e.g., Street < City < State < Country).
UNIT-III: Classification
Submission Date 20th April
Q1. Explain the Decision Tree Induction process. Define the three types of nodes (Root,
Internal, and Leaf) and explain the "stopping criteria" used to halt the growth of a tree.
Q2. A classifier for a "Rain" prediction model yields the following results:
True Positives (Predicted Rain, It Rained): 30
False Positives (Predicted Rain, It Stayed Dry): 10
True Negatives (Predicted Dry, It Stayed Dry): 50
False Negatives (Predicted Dry, It Rained): 10
Calculate: Accuracy, Precision, Recall, and the F1-Score.
Q3. Explain the K-fold Cross-Validation technique. If you have 1000 records and you
perform 5-fold Cross-Validation, how many records are used for training and how
many for testing in each iteration? Why is this better than a simple train/test split?
Q4. Explain the role of Linear Regression in data mining. Derive the method of Least
Squares to find the line of best fit ($Y = mx + c$). How is the accuracy of a regression
model measured using Mean Squared Error (MSE)?
Q5. Given the following dataset, use the Naive Bayes Classifier to predict if a game will be
played on a day where the conditions are:
X = ({Outlook=Rainy, Temperature=Cool, Humidity=High, Windy=True})
Day Outlook Temp Humidity Windy Play (Class)
D1 Sunny Hot High False No
D2 Sunny Hot High True No
D3 Overcast Hot High False Yes
D4 Rainy Mild High False Yes
D5 Rainy Cool Normal False Yes
Day Outlook Temp Humidity Windy Play (Class)
D6 Rainy Cool Normal True No
D7 Overcast Cool Normal True Yes
D8 Sunny Mild High False No
D9 Sunny Cool Normal False Yes
D10 Rainy Mild Normal False Yes
Steps to solve:
1. Calculate the Prior Probabilities P(Yes) and P(No)
2. Calculate the Conditional Probabilities for each attribute given the class (e.g.,
P(Rainy|Yes), P(Cool|Yes)).
3. Apply the Naive Bayes formula to find the posterior probability for both classes.
4. State the final classification.
UNIT-IV: ARM
Submission Date 20th April
Q1. Define Association Rule Mining. Explain the difference between Support and
Confidence, Lift with a simple example of a grocery transaction.
Q2. (The Apriori Principle) State the Apriori Property (also known as the downward
closure property). Why is this principle vital for improving the efficiency of frequent itemset
mining?
Q3. (Numerical: Apriori Algorithm) Given the following transaction database
(min_support = 2):
T1: {1, 2, 5}
T2: {2, 4}
T3: {2, 3}
T4: {1, 2, 4}
T5: {1, 3}
T6: {2, 3}
T7: {1, 3}
T8: {1, 2, 3, 5}
T9: {1, 2, 3} Find all frequent itemsets using the Apriori Algorithm.
From the frequent itemset found in Q3, generate all possible association rules. Minimum
confidence is 70%.
Q4. What are the major disadvantages of the Apriori algorithm? Explain how the FP-Growth
(Frequent Pattern Growth) algorithm overcomes these limitations.
Q5. Explain the Partitioning Algorithm for association rule mining. How does it manage to
find all frequent itemsets using only two passes over the database?
UNIT-V: Clustering & Advanced Mining
Submission Date 25th April
Q1. You have four 1D data points: 2, 4, 10, 12.
(a) If K=2 and the initial centroids are C_1 = 2 and C_2 = 10, show the clusters formed
after the first iteration.
(b) Calculate the new centroids for the second iteration.
(c) When does the K-Means algorithm stop?
Q2. Explain the difference between Agglomerative (Bottom-up) and Divisive (Top-down)
hierarchical clustering. Use a simple diagram of a Dendrogram to illustrate how clusters are
merged.
Q3. In hierarchical clustering, define how the distance between two clusters is calculated
using:
1. Single Linkage (Minimum distance)
2. Complete Linkage (Maximum distance)
3. Average Linkage
Q4. Why is Text Mining considered more difficult than traditional data mining? Briefly
explain the concept of TF-IDF (Term Frequency-Inverse Document Frequency) used to
identify important words in a document.
Q5. (a) What is Spatial Mining? Give an example related to GPS or satellite data.
(b) What are the unique challenges of Multimedia Data Mining (Images, Video, Audio)
compared to mining text-based tables?
(c) Define Temporal Data Mining. How does a Time-Series dataset differ from a standard
relational dataset? Provide an example involving stock market data or weather patterns.