2020 question paper
Data Mining
26 questions
Q1a. The full form of KDD is (i) knowledge database (ii) knowledge discovery in databases (iii) knowledge data division (iv) knowledge data definition20202m
Module 2: Data Mining and Association Rule Mining
View this question on its own page →The full form of KDD is
(i) knowledge database
(ii) knowledge discovery in databases
(iii) knowledge data division
(iv) knowledge data definitionWorked SolutionAnswer: (ii) knowledge discovery in databases
KDD stands for Knowledge Discovery in Databases — the overall, multi-step process of turning raw data into useful, actionable knowledge. Data mining is technically just one step within KDD (the pattern-extraction step), while KDD as a whole covers:
Data ─▶ Selection ─▶ Preprocessing ─▶ Transformation ─▶ Data Mining ─▶ Interpretation/Evaluation ─▶ KnowledgePeople often use "data mining" and "KDD" interchangeably in casual usage, but formally, data mining is embedded inside the broader KDD pipeline.
Q1b. You are given data about seismic activity in Japan, and you want to predict a magnitude of the next earthquake, this is an example of (i) supervised learning (ii) unsupervised learning (iii) serration (iv) dimensionality reduction20202m
Module 3: Classification and Prediction
View this question on its own page →You are given data about seismic activity in Japan, and you want to predict a magnitude of the next earthquake, this is an example of
(i) supervised learning
(ii) unsupervised learning
(iii) serration
(iv) dimensionality reductionQ1c. Which of the following does not involve in data mining? (i) Knowledge extraction (ii) Data archaeology (iii) Data exploration (iv) Data transformation20202m
Module 2: Data Mining and Association Rule Mining
View this question on its own page →Which of the following does not involve in data mining?
(i) Knowledge extraction
(ii) Data archaeology
(iii) Data exploration
(iv) Data transformationWorked SolutionAnswer: (iv) Data transformation
Data mining is historically also known by several near-synonyms:
- Knowledge extraction (extracting patterns/knowledge from data)
- Data archaeology (digging through historical data for insight)
- Data exploration (exploratory analysis of a dataset)
All three of these describe the act of mining itself. Data transformation, however, is a preprocessing step (converting/normalizing data into a suitable form) that happens before mining — it's not itself a synonym for or activity that defines data mining, making it the odd one out.
Q1d. _______ is a comparison of the general features of the target class data objects against the general features of objects from one or multiple contrasting classes. (i) Data characterization (ii) Data classification (iii) Data discrimination (iv) Data selection20202m
Module 2: Data Mining and Association Rule Mining
View this question on its own page →_______ is a comparison of the general features of the target class data objects against the general features of objects from one or multiple contrasting classes.
(i) Data characterization
(ii) Data classification
(iii) Data discrimination
(iv) Data selectionWorked SolutionAnswer: (iii) Data discrimination
- Data characterization summarizes the general features of a single target class (e.g., "typical profile of high-spending customers").
- Data discrimination compares the target class against one or more contrasting classes, highlighting the differences (e.g., "how do high-spending customers differ from low-spending customers?").
Since the question explicitly describes a comparison against contrasting classes, the answer is data discrimination, not characterization (which looks at only one class in isolation).
Q1e. Bayesian classifiers is (i) a class of learning algorithm that tries to find an optimum classification of a set of examples using the probabilistic theory (ii) any mechanism employed by a learning system to constrain the search space of a hypothesis (iii) an approach to the design of learning algorithms that is inspired by the fact that when people encounter new situations, they often explain them by reference to familiar experiences, adapting the explanations to fit the new situation (iv) None of the above20202m
Module 3: Classification and Prediction
View this question on its own page →Bayesian classifiers is
(i) a class of learning algorithm that tries to find an optimum classification of a set of examples using the probabilistic theory
(ii) any mechanism employed by a learning system to constrain the search space of a hypothesis
(iii) an approach to the design of learning algorithms that is inspired by the fact that when people encounter new situations, they often explain them by reference to familiar experiences, adapting the explanations to fit the new situation
(iv) None of the aboveQ1f. The output of KDD is (i) data (ii) information (iii) query (iv) useful information20202m
Module 2: Data Mining and Association Rule Mining
View this question on its own page →The output of KDD is
(i) data
(ii) information
(iii) query
(iv) useful informationWorked SolutionAnswer: (iv) useful information
KDD's whole purpose is to convert raw data into useful, actionable knowledge/information through a pipeline of selection, preprocessing, transformation, mining, and interpretation:
Raw Data ──KDD process──▶ Useful, actionable Information (patterns, rules, insights)Simply "data" (i) or a bare "query" (iii) is not the output of the process — those are inputs/tools. Generic "information" (ii) is too broad; the defining goal of KDD is specifically useful information that supports decision-making, which is why (iv) is the precise answer.
Q1g. Cluster is (i) group of similar objects that differ significantly from other objects (ii) operations on a database to transform or simplify data in order to prepare it for a machine learning algorithm (iii) symbolic representation of facts or ideas from which information can potentially be extracted (iv) None of the above20202m
Module 4: Cluster Analysis
View this question on its own page →Cluster is
(i) group of similar objects that differ significantly from other objects
(ii) operations on a database to transform or simplify data in order to prepare it for a machine learning algorithm
(iii) symbolic representation of facts or ideas from which information can potentially be extracted
(iv) None of the aboveQ1h. Background knowledge referred to (i) additional acquaintance used by a learning algorithm to facilitate the learning process (ii) a neural network that makes use of a hidden layer (iii) it is a form of automatic learning (iv) None of the above20202m
Module 2: Data Mining and Association Rule Mining
View this question on its own page →Background knowledge referred to
(i) additional acquaintance used by a learning algorithm to facilitate the learning process
(ii) a neural network that makes use of a hidden layer
(iii) it is a form of automatic learning
(iv) None of the aboveWorked SolutionAnswer: (i) additional acquaintance used by a learning algorithm to facilitate the learning process
Background knowledge refers to prior/domain knowledge (concept hierarchies, known constraints, expert rules, etc.) that is supplied to a learning/mining algorithm in addition to the raw training data, to guide and improve the learning process — e.g., helping generalize concepts, constrain hypothesis search, or interpret discovered patterns more meaningfully.
It is not itself a specific architecture like a hidden-layer neural network (ii), nor a synonym for "automatic learning" in general (iii) — it's specifically the extra contextual information fed into a learning system.
Q1i. Case-based learning is (i) a class of learning algorithm that tries to find an optimum classification of a set of examples using the probabilistic theory (ii) any mechanism employed by a learning system to constrain the search space of a hypothesis (iii) an approach to the design of learning algorithms that is inspired by the fact that when people encounter new situations, they often explain them by reference to familiar experiences, adapting the explanations to fit the new situation (iv) None of the above20202m
Module 3: Classification and Prediction
View this question on its own page →Case-based learning is
(i) a class of learning algorithm that tries to find an optimum classification of a set of examples using the probabilistic theory
(ii) any mechanism employed by a learning system to constrain the search space of a hypothesis
(iii) an approach to the design of learning algorithms that is inspired by the fact that when people encounter new situations, they often explain them by reference to familiar experiences, adapting the explanations to fit the new situation
(iv) None of the aboveQ1j. Some telecommunication companies want to segment their customers into distinct groups in order to send appropriate subscription offers. This is an example of (i) supervised learning (ii) data extraction (iii) serration (iv) unsupervised learning20202m
Module 4: Cluster Analysis
View this question on its own page →Some telecommunication companies want to segment their customers into distinct groups in order to send appropriate subscription offers. This is an example of
(i) supervised learning
(ii) data extraction
(iii) serration
(iv) unsupervised learningQ2a. Compare and contrast data warehouse system and operational database system.20207m
Module 1: Data Warehousing and Business Analysis
View this question on its own page →Compare and contrast data warehouse system and operational database system.
Worked SolutionData Warehouse vs Operational Database (OLTP)
Feature Operational DB (OLTP) Data Warehouse (OLAP) Purpose Day-to-day transaction processing Analysis, decision support Users Clerks, DBAs, front-line staff Managers, analysts, executives Data content Current, detailed data Historical, summarized, consolidated data Data model Normalized (E-R model), application-oriented Denormalized — star/snowflake schema, subject-oriented Access pattern Short, atomic transactions (read/write/update) Complex queries, mostly read-only, large scans Operations Insert, update, delete (many small txns) Aggregate, drill-down, roll-up (few large queries) Design goal High concurrency, fast single-record access Fast aggregation over large volumes Data volume per query Small Large Time orientation Reflects current value Reflects data over time (historical snapshots) Redundancy Minimized (normalization) Encouraged (denormalization for query speed) Backup/recovery Critical, transaction-consistent Less frequent, can be reloaded from source Conceptual Diagram
OLTP Systems Data Warehouse ┌───────────┐ ETL (Extract, ┌────────────────┐ │ Sales DB │──Transform,Load──▶│ │ ├───────────┤ │ Subject- │ │ HR DB │──────────────────▶│ oriented, │──▶ OLAP / Reports ├───────────┤ │ integrated, │ │ Inventory │──────────────────▶│ time-variant, │ │ DB │ │ non-volatile │ └───────────┘ └────────────────┘ (many small, (one large, normalized DBs) denormalized store)In short: OLTP is built for doing the business (fast writes, current data), a Data Warehouse is built for understanding the business (fast reads, historical/aggregated data). Bill Inmon defines the DW as subject-oriented, integrated, time-variant, and non-volatile — all properties an OLTP system deliberately does not have.
Q2b. Describe the steps involved in data mining when viewed as a process of knowledge discovery.20207m
Module 2: Data Mining and Association Rule Mining
View this question on its own page →Describe the steps involved in data mining when viewed as a process of knowledge discovery.
Worked SolutionData Mining as Knowledge Discovery — Steps
Data mining, viewed within the KDD (Knowledge Discovery in Databases) framework, involves the following iterative steps:
Raw Data │ ▼ 1. Data Cleaning ───────────▶ remove noise, inconsistent data │ ▼ 2. Data Integration ────────▶ combine multiple data sources │ ▼ 3. Data Selection ──────────▶ retrieve data relevant to the task │ ▼ 4. Data Transformation ─────▶ consolidate into forms suitable for mining │ (aggregation, normalization) ▼ 5. Data Mining ─────────────▶ apply algorithms to extract patterns │ (classification, clustering, association) ▼ 6. Pattern Evaluation ──────▶ identify truly interesting patterns │ using interestingness measures ▼ 7. Knowledge Presentation ──▶ visualize/present mined knowledge │ (reports, graphs, rules) ▼ Useful KnowledgeStep-by-step explanation
- Data Cleaning — removes noise and inconsistent/incomplete data.
- Data Integration — merges data from multiple heterogeneous sources into one coherent store.
- Data Selection — retrieves only the data relevant to the analysis task at hand.
- Data Transformation — converts/consolidates data into forms appropriate for mining (e.g., aggregation, normalization, discretization).
- Data Mining — the essential step where intelligent methods (classification, clustering, association rule mining, etc.) are applied to extract data patterns.
- Pattern Evaluation — identifies truly interesting patterns representing knowledge, based on interestingness measures (support, confidence, novelty).
- Knowledge Presentation — visualization and knowledge representation techniques present the mined knowledge to users in an understandable form.
Steps 1–4 together are often called preprocessing, and this whole pipeline is typically iterative — insights from later steps may send you back to refine earlier ones (e.g., pattern evaluation might reveal that different data selection is needed).
Q3a. What is data warehouse backend process? Explain briefly.20207m
Module 1: Data Warehousing and Business Analysis
View this question on its own page →What is data warehouse backend process? Explain briefly.
Worked SolutionData Warehouse Backend Process
The backend process is the pipeline that moves data from scattered operational sources into the clean, integrated warehouse. It runs "behind the scenes," invisible to end users, and is commonly called ETL (Extract–Transform–Load).
Operational ┌─────────┐ ┌───────────┐ ┌────────┐ ┌───────────┐ Source Systems ─▶│ Extract │─────▶│ Clean/ │─────▶│Transform│─────▶│ Load │──▶ Data (RDBMS, files, │ │ │ Scrub │ │ │ │ (+Refresh)│ Warehouse ERP, legacy) └─────────┘ └───────────┘ └────────┘ └───────────┘Steps
- Extraction — Data is pulled from multiple heterogeneous source systems (databases, flat files, ERP/CRM, external feeds).
- Cleaning (Scrubbing) — Detects and corrects errors, inconsistencies, missing values, and duplicate records to improve data quality.
- Transformation — Converts data into the warehouse's standard format: normalizing units, renaming fields, applying business rules, aggregating, and generating derived/summary values.
- Integration — Merges data from different sources into a unified, subject-oriented schema, resolving naming and structural conflicts (schema/semantic integration).
- Loading — Writes the cleaned, transformed, integrated data into the warehouse's fact and dimension tables, often in bulk/batch mode.
- Refreshing — Periodically propagates updates from the source systems into the warehouse (full refresh or incremental refresh) to keep it current.
Why it matters
The backend process is what actually enforces the warehouse's defining properties — integrated (step 4), non-volatile (loaded once, not transaction-updated), and time-variant (refresh keeps historical snapshots) — turning raw, messy operational data into a trustworthy analytical resource.
Q3b. Explain different OLAP operations.20207m
Module 1: Data Warehousing and Business Analysis
View this question on its own page →Explain different OLAP operations.
Worked SolutionOLAP Operations
OLAP tools let users navigate a multidimensional data cube through five core operations:
1. Roll-up (Drill-up)
Aggregates data by climbing up a concept hierarchy or reducing dimensions.
Day → Month → Quarter → Year (climbing the Time hierarchy)Example: Sales by City → Sales by State → Sales by Country.
2. Drill-down
The reverse of roll-up — moves down the hierarchy to reveal finer detail.
Year → Quarter → Month → DayExample: Yearly sales → Quarterly sales → Monthly sales.
3. Slice
Selects a single value on one dimension, producing a lower-dimensional sub-cube.
Cube(Time, Product, Region) --Time="2024"--> Table(Product, Region)4. Dice
Selects a sub-cube by specifying ranges/values on two or more dimensions.
Dice: Time∈{2023,2024}, Region="West" → smaller cube5. Pivot (Rotate)
Rotates the data axes to provide an alternate presentation — e.g., swapping rows and columns in a cross-tab.
Before: After pivot: Q1 Q2 Q3 Prod A Prod B Prod A 10 20 15 ──▶ Q1 10 12 Prod B 12 18 22 Q2 20 18 Q3 15 22Summary Table
Operation Effect Dimensionality change Roll-up Aggregate, coarsen granularity Decreases (or stays, fewer dims) Drill-down Disaggregate, finer granularity Increases Slice Fix one dimension's value Decreases by 1 Dice Restrict 2+ dimensions Same dims, smaller sub-cube Pivot Re-orient axes for viewing Unchanged These operations give analysts an interactive way to explore a cube from summary to detail and back, without re-running queries against the raw source data.
Q4a. What is data cleaning? Describe various approaches for cleaning data having missing values.20205m
Module 2: Data Mining and Association Rule Mining
View this question on its own page →What is data cleaning? Describe various approaches for cleaning data having missing values.
Worked SolutionData Cleaning & Approaches for Missing Values
What is Data Cleaning?
Data cleaning (data cleansing/scrubbing) is the preprocessing step that detects and corrects (or removes) inaccurate, incomplete, inconsistent, or noisy data from a dataset — filling missing values, smoothing noise, identifying/handling outliers, and resolving inconsistencies — so that mining algorithms operate on trustworthy, high-quality data.
Approaches for Handling Missing Values
Ignore the tuple
Discard the entire record. Only reasonable when the missing attribute is essential (e.g., the class label) and the dataset is large enough to absorb the loss.Fill in manually
A domain expert inspects and enters the correct value. Accurate but time-consuming and infeasible for large datasets.Use a global constant
Replace all missing entries with a fixed placeholder (e.g.,"Unknown",-999). Simple but risky — the mining algorithm may misinterpret the placeholder as a meaningful pattern.Use the attribute mean/median
Fill numeric gaps with the mean (or median, for skewed distributions) of that attribute computed over all available tuples.Use the mean/median for the same class
More accurate than a global mean — uses the mean of tuples belonging to the same class as the one with the missing value.Use the most probable value
Predict the missing value using inference-based methods: regression, Bayesian formalism, or decision-tree induction based on other attributes. This is the most sophisticated and generally most accurate method, since it uses the relationships already present in the data.
Missing Value Handling │ ┌─────────┬──────────┼──────────┬────────────┬───────────┐ Ignore Manual Global Attribute Class-wise Predicted tuple fill constant mean/median mean/median (regression/ Bayes/tree)In practice, methods 4–6 (statistical/model-based imputation) are most commonly used in real mining pipelines because they preserve dataset size while introducing minimal bias, compared to simply dropping records.
Q4b. Use the two methods below to normalize the following group of data: 200; 300; 400; 600; 1000 * min-max normalization by setting min = 0 and max = 1 * z-score normalization20205m
Module 2: Data Mining and Association Rule Mining
View this question on its own page →Use the two methods below to normalize the following group of data:
200; 300; 400; 600; 1000
- min-max normalization by setting min = 0 and max = 1
- z-score normalization
Worked SolutionNormalizing Data: 200, 300, 400, 600, 1000
Method 1: Min-Max Normalization (new_min=0, new_max=1)
Here , , range , new_min=0, new_max=1, so simply:
v Calculation v' 200 (200-200)/800 0.000 300 (300-200)/800 0.125 400 (400-200)/800 0.250 600 (600-200)/800 0.500 1000 (1000-200)/800 1.000 Method 2: Z-Score Normalization
Mean:
Variance (population):
Std dev:
v v − μ z = (v−μ)/σ 200 −300 −1.06 300 −200 −0.71 400 −100 −0.35 600 +100 +0.35 1000 +500 +1.77 Summary
Original Min-Max [0,1] Z-Score 200 0.000 −1.06 300 0.125 −0.71 400 0.250 −0.35 600 0.500 +0.35 1000 1.000 +1.77 Min-max squeezes everything cleanly into [0,1] (bounded by the observed extremes), while z-score centers the data at 0 with unit spread — the value 1000, being furthest from the mean, gets the largest magnitude z-score (+1.77), reflecting how much of an outlier it is relative to the rest of the group.
Q4c. What is the value range of the following normalization methods? * min-max normalization * z-score normalization * normalization by decimal scaling20204m
Module 2: Data Mining and Association Rule Mining
View this question on its own page →What is the value range of the following normalization methods?
- min-max normalization
- z-score normalization
- normalization by decimal scaling
Worked SolutionValue Ranges of Normalization Methods
Method Formula Value Range Min-Max Normalization Bounded exactly to [new_min, new_max] (commonly [0,1]) Z-Score Normalization Unbounded in theory; in practice typically falls roughly in [−3, +3] for approximately normal data Decimal Scaling , where is the smallest integer such that Bounded to (−1, 1) Notes
- Min-max preserves the exact relationships in the original data and gives a strictly bounded range — but it's sensitive to outliers (a single extreme value stretches/compresses everything else).
- Z-score has no hard bounds since it depends on how far a value is from the mean in standard-deviation units — extreme outliers can produce large |z| values well beyond ±3.
- Decimal scaling normalizes by moving the decimal point, controlled by the maximum absolute value in the data, guaranteeing the result always lies strictly between −1 and 1.
Min-Max: [0 ──────────── 1] (fixed, bounded) Z-Score: (-∞ ... -3 0 +3 ... +∞) (unbounded, ~[-3,3] typical) Decimal Scale: (-1 ──────── 1) (bounded, data-dependent)Q5a. Write and explain pseudocode for a priori algorithm. Explain the terms: (i) support count; (ii) confidence.20206m
Module 2: Data Mining and Association Rule Mining
View this question on its own page →Write and explain pseudocode for a priori algorithm. Explain the terms:
(i) support count;
(ii) confidence.Worked SolutionApriori Algorithm — Pseudocode & Key Terms
Pseudocode
Algorithm Apriori(D, min_sup): Input: D = transaction database, min_sup = minimum support threshold Output: L = all frequent itemsets in D L1 = {frequent 1-itemsets found by scanning D} k = 2 while (L(k-1) is not empty): Ck = apriori_gen(L(k-1)) // candidate generation for each transaction t in D: Ct = subset(Ck, t) // candidates contained in t for each candidate c in Ct: c.count++ Lk = { c in Ck | c.count >= min_sup } L = L ∪ Lk k = k + 1 return L Function apriori_gen(L(k-1)): Ck = ∅ // Join step: join L(k-1) with itself for each pair (p, q) in L(k-1) x L(k-1): if p and q share the first (k-2) items: c = p ∪ {last item of q} // Prune step: remove candidates with an infrequent subset if all (k-1)-subsets of c are in L(k-1): Ck = Ck ∪ {c} return CkExplanation
- First pass: count individual items to find frequent 1-itemsets (L1).
- Iterative passes: for k = 2, 3, ..., generate candidate k-itemsets (Ck) by joining frequent (k-1)-itemsets that share a common (k-2)-item prefix.
- Prune step: discard any candidate whose any (k-1)-subset is not frequent (using the apriori property: all subsets of a frequent itemset must be frequent).
- Count & filter: scan the database, count support for surviving candidates, keep only those meeting min_sup → Lk.
- Repeat until no new frequent itemsets are found.
Key Terms
(i) Support Count
The number of transactions in the database that contain a given itemset:
Support (as a fraction) = support_count / |D|. It measures how frequently an itemset appears.(ii) Confidence
For a rule , confidence measures how reliably B follows when A is present:
It is the conditional probability — the fraction of transactions containing A that also contain B.Support: "How often does this itemset occur overall?" Confidence: "Given the antecedent occurred, how often does the consequent follow?"Q5b. Consider a database D, consisting of 15 transactions. Suppose minimum support count is 2 (i.e., min_sup = 20%) and minimum confidence required is 70%. Find out the frequent item set using a priori algorithm. Explain each step with diagram: | A1 | A2 | A3 | A4 | A5 | A6 | A7 | A8 | A9 | |:---:|:---:|:---:|:---:|:---:|:---:|:---:|:---:|:---:| | 1 | 0 | 0 | 0 | 1 | 1 | 0 | 1 | 0 | | 0 | 1 | 0 | 1 | 0 | 0 | 0 | 1 | 0 | | 0 | 0 | 0 | 1 | 1 | 0 | 1 | 0 | 0 | | 0 | 1 | 1 | 0 | 0 | 0 | 0 | 0 | 0 | | 0 | 0 | 0 | 0 | 1 | 1 | 1 | 0 | 0 | | 0 | 1 | 1 | 1 | 0 | 0 | 0 | 0 | 0 | | 0 | 1 | 0 | 0 | 0 | 1 | 1 | 0 | 1 | | 0 | 0 | 0 | 0 | 1 | 0 | 0 | 0 | 0 | | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 1 | 0 | | 0 | 0 | 1 | 0 | 1 | 0 | 1 | 0 | 0 | | 0 | 0 | 1 | 0 | 1 | 0 | 1 | 0 | 0 | | 0 | 0 | 0 | 0 | 1 | 1 | 0 | 1 | 0 | | 0 | 1 | 0 | 1 | 0 | 1 | 1 | 0 | 0 | | 1 | 0 | 1 | 0 | 1 | 0 | 1 | 0 | 0 | | 0 | 1 | 1 | 0 | 0 | 0 | 0 | 0 | 1 |20208m
Module 2: Data Mining and Association Rule Mining
View this question on its own page →Consider a database D, consisting of 15 transactions. Suppose minimum support count is 2 (i.e., min_sup = 20%) and minimum confidence required is 70%. Find out the frequent item set using a priori algorithm. Explain each step with diagram:
A1 A2 A3 A4 A5 A6 A7 A8 A9 1 0 0 0 1 1 0 1 0 0 1 0 1 0 0 0 1 0 0 0 0 1 1 0 1 0 0 0 1 1 0 0 0 0 0 0 0 0 0 0 1 1 1 0 0 0 1 1 1 0 0 0 0 0 0 1 0 0 0 1 1 0 1 0 0 0 0 1 0 0 0 0 0 0 0 0 0 0 0 1 0 0 0 1 0 1 0 1 0 0 0 0 1 0 1 0 1 0 0 0 0 0 0 1 1 0 1 0 0 1 0 1 0 1 1 0 0 1 0 1 0 1 0 1 0 0 0 1 1 0 0 0 0 0 1 Worked SolutionApriori Algorithm — 15-Transaction Database (9 Binary Attributes)
Given: min_sup count = 2, min_conf = 70%, database D of 15 transactions over items A1–A9.
Step 1: Scan D → Count 1-itemsets (C1)
Item Support count Frequent (≥2)? A1 2 ✅ A2 6 ✅ A3 6 ✅ A4 4 ✅ A5 8 ✅ A6 5 ✅ A7 7 ✅ A8 4 ✅ A9 2 ✅ L1 = {A1, A2, A3, A4, A5, A6, A7, A8, A9} — every item clears the low min_sup=2 threshold.
L1: [A1]─[A2]─[A3]─[A4]─[A5]─[A6]─[A7]─[A8]─[A9] 2 6 6 4 8 5 7 4 2Step 2: Generate C2 (all pairs of L1), scan D, prune → L2
Only pairs with support count ≥ 2 survive (out of the 36 possible pairs, most drop out):
Pair Count Pair Count A1,A5 2 ✅ A4,A7 2 ✅ A2,A3 3 ✅ A5,A6 3 ✅ A2,A4 3 ✅ A5,A7 5 ✅ A2,A6 2 ✅ A5,A8 2 ✅ A2,A7 2 ✅ A6,A7 3 ✅ A2,A9 2 ✅ A6,A8 2 ✅ A3,A5 3 ✅ A3,A7 3 ✅ (All other 22 pairs, e.g. A1-A2, A1-A3, A2-A5, A3-A4, A4-A5, A7-A8, etc., had count ≤1 and were pruned.)
L2 = {A1A5, A2A3, A2A4, A2A6, A2A7, A2A9, A3A5, A3A7, A4A7, A5A6, A5A7, A5A8, A6A7, A6A8} (14 itemsets)
L1 (9 items) ──join+prune──▶ L2 (14 pairs, out of 36 possible)Step 3: Generate C3 via apriori-gen (join L2 with itself + prune using apriori property)
Joining pairs sharing a common first item, keeping only candidates whose every 2-subset is in L2:
Candidate 2-subsets check Valid? DB count Frequent? A2,A3,A7 A2A3✅ A2A7✅ A3A7✅ ✅ 0 ❌ A2,A4,A7 A2A4✅ A2A7✅ A4A7✅ ✅ 1 ❌ A2,A6,A7 A2A6✅ A2A7✅ A6A7✅ ✅ 2 ✅ A3,A5,A7 A3A5✅ A3A7✅ A5A7✅ ✅ 3 ✅ A5,A6,A7 A5A6✅ A5A7✅ A6A7✅ ✅ 1 ❌ A5,A6,A8 A5A6✅ A5A8✅ A6A8✅ ✅ 2 ✅ L3 = {A2A6A7(2), A3A5A7(3), A5A6A8(2)}
Step 4: Try C4
No two itemsets in L3 share a common 2-item prefix (A2A6A7, A3A5A7, A5A6A8 are all disjoint in their first two items) → no valid C4 candidates → algorithm terminates.
Final Frequent Itemsets
L1 (9): A1 A2 A3 A4 A5 A6 A7 A8 A9 L2 (14): A1A5 A2A3 A2A4 A2A6 A2A7 A2A9 A3A5 A3A7 A4A7 A5A6 A5A7 A5A8 A6A7 A6A8 L3 (3): A2A6A7 A3A5A7 A5A6A8Total = 26 frequent itemsets
Strong Rules Example (from the maximal itemset A3,A5,A7 — count 3)
Confidence check for rules generated from {A3,A5,A7}:
- A3,A5 ⇒ A7: conf = 3/support(A3A5)=3/3 = 100% ✅ strong
- A3,A7 ⇒ A5: conf = 3/support(A3A7)=3/3 = 100% ✅ strong
- A5,A7 ⇒ A3: conf = 3/support(A5A7)=3/5 = 60% ❌ below 70% threshold
Only the first two rules meet min_conf = 70% and would be output as strong association rules.
Q6. Draw decision tree for the following data sets. Use entropy as a node selection mechanism: | Outlook | Temp (F) | Humidity | Windy | Class | |:---|:---|:---|:---|:---| | Rainy | Hot | High | False | No | | Rainy | Hot | High | True | No | | Overcast | Hot | High | False | Yes | | Sunny | Mild | High | False | Yes | | Sunny | Cool | Normal | False | Yes | | Sunny | Cool | Normal | True | No | | Overcast | Cool | Normal | True | Yes | | Rainy | Mild | High | False | No | | Rainy | Cool | Normal | False | Yes | | Sunny | Mild | Normal | False | Yes | | Rainy | Mild | Normal | True | Yes | | Overcast | Mild | High | True | Yes | | Overcast | Hot | Normal | False | Yes | | Sunny | Mild | High | True | No |202014m
Module 3: Classification and Prediction
View this question on its own page →Draw decision tree for the following data sets. Use entropy as a node selection mechanism:
Outlook Temp (F) Humidity Windy Class Rainy Hot High False No Rainy Hot High True No Overcast Hot High False Yes Sunny Mild High False Yes Sunny Cool Normal False Yes Sunny Cool Normal True No Overcast Cool Normal True Yes Rainy Mild High False No Rainy Cool Normal False Yes Sunny Mild Normal False Yes Rainy Mild Normal True Yes Overcast Mild High True Yes Overcast Hot Normal False Yes Sunny Mild High True No Q7a. What are the main requirements for cluster analysis?20207m
Module 4: Cluster Analysis
View this question on its own page →What are the main requirements for cluster analysis?
Q7b. Explain different basic clustering methods.20207m
Module 4: Cluster Analysis
View this question on its own page →Explain different basic clustering methods.
Q8a. What is multilevel association rule mining? Explain different approaches to do multilevel association rule mining.20207m
Module 2: Data Mining and Association Rule Mining
View this question on its own page →What is multilevel association rule mining? Explain different approaches to do multilevel association rule mining.
Worked SolutionMultilevel Association Rule Mining
What is it?
Multilevel association rule mining discovers association rules at multiple levels of abstraction, using a concept hierarchy over the items (e.g., Item → Subcategory → Category). Mining only at the lowest (raw item) level often misses valuable patterns, because individual items may be too specific to have high support, while their generalized categories do.
All Items / \ Beverages Bakery / \ / \ Milk Juice Bread Cookies / \ Amul Milk Nestle MilkRules can be mined at any level:
- Low level (too specific, low support): {Amul Milk} ⇒ {Nestle Cookies}
- High level (general, higher support): {Beverages} ⇒ {Bakery}
Approaches to Multilevel Mining
1. Uniform Minimum Support
The same min_sup threshold is used across all levels of the hierarchy.Level 1 (General): min_sup = 5% ──▶ applied Level 2 (Specific): min_sup = 5% ──▶ same threshold- Pros: simple, single threshold to set.
- Cons: high-level (general) items naturally have more support, while low-level (specific) items may fail the same threshold even if they represent genuinely important niche patterns — some interesting low-level rules are missed.
2. Reduced (Level-Specific) Minimum Support
Each level of the hierarchy has its own, typically lower, minimum support threshold as you go deeper.Level 1 (General): min_sup = 5% Level 2 (Category): min_sup = 3% Level 3 (Specific): min_sup = 1%- Pros: accounts for the fact that specific items are naturally rarer; lets genuinely interesting low-level patterns surface.
- Cons: more thresholds to tune; poor tuning can either flood results with noise or still miss patterns.
Within reduced support, there are two search strategies:
- Level-by-level independent — each level is mined independently regardless of results at other levels.
- Level-cross filtering by single/k-itemset — an itemset at a lower level is examined only if its corresponding ancestor itemset at the level above was found frequent (this prunes the search space significantly using apriori-style reasoning across levels).
3. Group-Based (Item-Specific) Minimum Support
Different groups of items (rather than levels) get their own custom support thresholds, letting users assign lower thresholds to specific item groups they know to be rare-but-important (e.g., luxury goods) and higher thresholds elsewhere.Summary
Approach Threshold granularity Trade-off Uniform min_sup One threshold, all levels Simple, but biased toward general-level rules Reduced min_sup One threshold per level Captures specific-level rules, more tuning needed Group-based min_sup Custom threshold per item group Most flexible, most tuning effort In short: multilevel mining lets analysts see both broad, general trends and narrow, specific patterns by mining across a concept hierarchy — with reduced/level-specific support being the most common practical approach to avoid missing important niche rules.
Q8b. Explain Naive Bayesian classification algorithm.20207m
Module 3: Classification and Prediction
View this question on its own page →Explain Naive Bayesian classification algorithm.
Q9a. With neat diagram, explain the architecture of data warehouse. Explain the terms ROLAP, MOLAP and HOLAP.20207m
Module 1: Data Warehousing and Business Analysis
View this question on its own page →With neat diagram, explain the architecture of data warehouse. Explain the terms ROLAP, MOLAP and HOLAP.
Worked SolutionData Warehouse Architecture (with diagram) + ROLAP, MOLAP, HOLAP
Architecture Diagram
┌───────────┐ ┌───────────┐ ┌───────────┐ │ Source DB1│ │ Source DB2│ │ Ext. Files│ └─────┬─────┘ └─────┬─────┘ └─────┬─────┘ └──────────────┼──────────────┘ ▼ ┌───────────────────┐ │ ETL: Extract, │ │ Clean, Transform, │ │ Load │ └──────────┬──────────┘ ▼ ┌───────────────────┐ │ Data Warehouse │◀──▶ Metadata Repository │ (fact + dimension │ │ tables) │ └──────────┬──────────┘ │ ┌───────────────┼───────────────┐ ▼ ▼ ▼ ┌───────────┐ ┌────────────┐ ┌────────────┐ │ Data Marts │ │ OLAP Server │ │ Mining │ │ (dept-wise)│ │(ROLAP/MOLAP/│ │ Engine │ │ │ │ HOLAP) │ │ │ └───────────┘ └────────────┘ └────────────┘ │ ▼ Front-end Tools (Query, Reporting, Dashboards)Layers
- Source systems — operational databases, files, external feeds.
- ETL layer — extracts, cleans, transforms, and loads data into the warehouse.
- Metadata repository — describes schema, source mappings, and lineage.
- Central warehouse — stores integrated, subject-oriented, time-variant data as fact/dimension tables.
- Data marts / OLAP servers / mining engine — provide focused or multidimensional access.
- Front-end tools — dashboards, reports, ad-hoc query tools for end users.
ROLAP, MOLAP, HOLAP (OLAP server types)
Type Storage Speed Scalability ROLAP (Relational OLAP) Star/snowflake tables in an RDBMS; multidimensional queries translated to SQL Slower (query-time computation) Very high — handles huge data volumes MOLAP (Multidimensional OLAP) Pre-computed proprietary cube (array structure) Fastest (pre-aggregated lookups) Limited — cube can explode in size for sparse/high-cardinality data HOLAP (Hybrid OLAP) Detail data in relational tables, summary data in cubes Balanced High — combines both strengths In short: the architecture moves data from many raw sources through ETL into one integrated warehouse, which is then served up via an OLAP layer — and ROLAP/MOLAP/HOLAP are simply three different engineering trade-offs (storage vs speed vs scale) for implementing that OLAP layer.
Q9b. What are the differences between the three main types of data warehouse usage—information processing, analytical processing and data mining? Discuss the motivation behind OLAP mining (OLAM).20207m
Module 1: Data Warehousing and Business Analysis
View this question on its own page →What are the differences between the three main types of data warehouse usage—information processing, analytical processing and data mining? Discuss the motivation behind OLAP mining (OLAM).
Worked SolutionData Warehouse Usage Types & Motivation for OLAM
(Same underlying question as the 2022 paper — see below for the full breakdown.)
Three Types of Data Warehouse Usage
Aspect Information Processing Analytical Processing Data Mining Goal Query, basic statistics, reporting Multidimensional analysis (OLAP) Discover hidden patterns/knowledge Operations Simple queries, cross-tabs, charts Roll-up, drill-down, slice, dice, pivot Classification, clustering, association, prediction User interaction Passive — user asks, system reports Interactive — user explores the cube Can be automatic — system discovers patterns Output Reports, summary tables Multidimensional views Rules, models, clusters, predictions Example "Total sales last quarter" "Sales by Region × Product × Time, drilled to month" "Which customers are likely to churn?" - Information processing answers known questions with simple queries/reports.
- Analytical processing (OLAP) lets users interactively explore a multidimensional cube.
- Data mining automatically discovers patterns the user didn't know to ask about.
Motivation for OLAP Mining (OLAM)
OLAM integrates OLAP with data mining, applying mining directly on multidimensional cubes instead of raw flat data:
- High-quality warehouse data — already cleaned/integrated, giving more reliable mining results than mining raw operational data.
- Reusable infrastructure — ETL, metadata, and OLAP engines built for the warehouse serve mining too, avoiding duplicated effort.
- OLAP-based exploratory analysis — analysts drill/roll/slice/dice to narrow down interesting regions before mining, far more efficient than mining an entire raw dataset.
- Flexible, on-line mining function selection — users can switch between classification, association, clustering, etc. at different cube cells/granularities within the same interactive session.
Data Warehouse (Cube) │ ┌─────────┼─────────┐ ▼ ▼ ▼ OLAP OLAM Mining (explore) (integrated) (patterns) └─────────┴─────────┘In short: OLAM combines OLAP's interactive navigation with mining's automatic pattern discovery, letting analysts first zoom into the relevant cube region, then mine it for deeper insight.