(SEM V) THEORY EXAMINATION 2024-25 INTRODUCTION TO DATA ANALYTICS AND VISUALIZATION

Course: B.Tech (Semester V)
Subject: Introduction to Data Analytics and Visualization
Subject Code: BCDS501
Exam Year: 2024–25
Maximum Marks: 70
Time: 3 Hours
Paper ID: 310469

Exam Structure Overview

SECTION A (2 × 7 = 14 Marks)
All short conceptual questions — definition-based or comparison-oriented.

Topics include scaling analytics, modern BI tools, Bayesian vs Neural networks, real-time sentiment analysis, clustering techniques (CLIQUE, ProCLUS), and visualization types.

SECTION B (3 × 7 = 21 Marks)
Attempt any three — descriptive questions requiring explanation and reasoning.

Topics: Big Data characteristics, rule induction, decaying windows, clustering comparisons, and visual interface design.

SECTION C (3 × 7 = 21 Marks)
Attempt one part from each question — application and analytical-type questions.

Covers IoT-driven analytics, analytic lifecycle, multivariate analysis, optimization methods, real-time analytics platforms, streaming clustering, and visualization techniques.

Key Topics for Preparation (From the Paper)

Big Data & Analytics Scaling

Volume, Velocity, Variety, Veracity, Value

Challenges: storage, processing speed, integration

Solutions: Hadoop, Spark, cloud analytics, data lakes

Modern BI Tools

Tableau, Power BI, Looker, Qlik Sense

Integration of data extraction, visualization, and reporting

Modeling Approaches

Bayesian models – probabilistic, interpretable, uncertainty estimates

Neural networks – deep learning, pattern recognition

Prediction Dynamics

Linear models (regression) vs Nonlinear (trees, neural nets)

Real-Time Sentiment Analysis

Social media data, NLP challenges, noise, slang, emojis

Clustering Techniques

CLIQUE / ProCLUS – subspace clustering for high-dimensional data

K-means vs Hierarchical – speed vs interpretability

Visualization Trade-offs

Static (reports, print) vs Dynamic (dashboards, interactivity)

Rule Induction

Decision trees, association rules (Apriori), pattern extraction

Decaying Windows (Stream Analytics)

Focus on recent data → efficient memory usage

Figure Captions in Visualization

Explain chart meaning, context, and message clarity

IoT & Real-Time Systems

Continuous sensor data → demand for real-time analytics

Analytic Lifecycle

Stages: data collection → preparation → modeling → evaluation → deployment

Multivariate & High-Dimensional Challenges

Dimensionality reduction (PCA, feature selection)

Optimization Techniques

Deterministic (gradient descent) vs Stochastic (genetic algorithms)

Real-Time Analytics Platforms (RTAP)

Examples: Kafka + Spark + Tableau; case study: stock prediction

Parallelism & Streaming Clustering

Distributed computing for large datasets; micro-clustering

Visualization Design

Navigation links, user interaction, human vision limits