Data Science Syllabus

The following syllabus map represents common subject families. Universities arrange subjects differently and may use different names.

Engineering Mathematics

Students study calculus, differential equations, matrices, probability, statistics, numerical methods and discrete Mathematics. These topics support algorithms, machine learning, graphics, networks and system modelling.

Discrete Mathematics is especially relevant because computing uses logic, sets, relations, graphs, combinatorics and proof. It strengthens rigorous reasoning beyond numerical calculation.

Programming fundamentals

Introductory programming teaches variables, control structures, functions, arrays, pointers or references, input-output and debugging. C is common because it reveals memory and system behaviour, while Python may be used for accessible problem-solving.

The objective is not to memorise syntax. Students should learn to break a problem into steps, select data representations, test edge cases and explain the result.

Object-oriented programming

Object-oriented programming introduces classes, objects, encapsulation, inheritance, polymorphism, exceptions and reusable design. Java, C++ or another language may be used. Laboratory projects often include small desktop, web or console applications.

Data structures

Data Structures covers arrays, linked lists, stacks, queues, trees, graphs, hashing and heaps. Students analyse how organisation affects search, insertion, deletion and memory use.

This is a core placement subject. Strong learning requires implementing structures and solving problems rather than reading finished code.

Design and analysis of algorithms

Algorithms examines complexity, recursion, sorting, searching, divide-and-conquer, greedy methods, dynamic programming, graph algorithms and computational limits. Students learn to evaluate solutions using time and space complexity.

Probability and statistical inference

Students study random variables, probability distributions, sampling, estimation, confidence intervals and hypothesis testing. These ideas help them separate real evidence from chance variation and describe uncertainty honestly.

Linear algebra and optimisation

Vectors, matrices, eigenvalues, decompositions, gradients and optimisation support regression, dimensionality reduction and machine learning. Exercises should connect the Mathematics with code and model behaviour.

Exploratory data analysis

Exploratory analysis examines data types, distributions, missing values, outliers and relationships before modelling. Students learn to investigate why an unusual value exists instead of deleting it automatically.

Data cleaning and preparation

Real datasets contain inconsistent labels, duplicates, invalid values and changing definitions. Students develop reproducible methods for validation, transformation, joining, encoding and documenting preparation decisions.

Data visualisation and storytelling

Students select charts according to the question and data type, design dashboards and explain findings to technical and non-technical audiences. They also learn how unsuitable scales, colours and selective reporting can mislead.

Statistical modelling

Regression, classification, analysis of variance and related methods help students connect variables and test assumptions. Interpretation, residual analysis and uncertainty deserve as much attention as predictive performance.

Database management systems

DBMS teaches data modelling, relational design, SQL, normalisation, transactions, concurrency, indexing and recovery. Students should learn both query writing and why a database preserves consistency.

Projects may include a web or mobile application backed by a relational database. Advanced electives may introduce distributed, NoSQL or data-warehouse systems.

Model evaluation and experimentation

Students learn train-validation-test separation, cross-validation and metrics such as precision, recall, F1 score and error measures. They examine class imbalance, leakage, overfitting, experiment design and A/B testing.

Software engineering

Software Engineering introduces requirements, design, development methods, testing, configuration management, maintenance and project planning. It shows why building production software involves more than writing code.

Team projects should use version control, issue tracking, review and documentation. These practices closely resemble professional work.

Big data and distributed processing

This area introduces distributed storage, batch and stream processing, scalable computation and large-data trade-offs. Architecture and problem selection matter more than memorising one platform's commands.

Time-series analysis

Time-series subjects cover trend, seasonality, autocorrelation, forecasting, back-testing and temporal leakage. Applications include demand, finance, sensors, energy and operations.

Business intelligence and dashboards

Students learn dimensional modelling, key performance indicators, reporting, interactive dashboards and stakeholder requirements. A visually attractive dashboard is useful only when definitions and source data are trustworthy.

Data privacy, ethics and governance

Students examine consent, purpose limitation, access control, anonymisation, bias, fairness, explainability and accountability. Legally accessible data is not automatically ethical or appropriate to use.

Artificial intelligence and machine learning

AI courses introduce search, reasoning, knowledge representation and intelligent agents. Machine learning may cover regression, classification, clustering, model evaluation and neural networks. Mathematics, clean data and programming are prerequisites for responsible work.

Data engineering and cloud systems

Students study ingestion, ETL and ELT pipelines, warehouses, lakes, orchestration, distributed processing and cloud services. Reliability, lineage, cost, security and recoverability are central concerns.

Natural language processing and computer vision

Advanced electives may cover text representation, language models, image features and neural networks. Students should understand dataset bias, computational cost and application-specific evaluation.

Deployment and MLOps

Deployment connects models with applications through APIs, containers, versioning and automated workflows. Monitoring covers drift, latency, failures and retraining. MLOps is normally an advanced pathway rather than an automatic fresher role.

Typical semester pattern

Laboratories

Important laboratory areas include Python or R, SQL, statistics, data cleaning, visualisation, machine learning, big-data processing, data engineering and project development.

A laboratory should require students to design, implement, test and explain work. Copying identical programs prevents skill development and creates difficulty during placements.

Electives

Common electives include cloud analytics, machine learning, business intelligence, computer vision, natural language processing, deep learning, time series, causal inference, distributed systems and optimisation.

Students should choose electives that form a coherent pathway. Selecting unrelated fashionable subjects without prerequisites can create shallow knowledge.

Project ideas

• environmental or agricultural data analysis;
• secure campus service application;
• retail demand forecasting system;
• public transport or mobility visualisation tool;
• smart energy controller;
• operating-system scheduling simulator;
• FPGA-based digital module;
• assistive computing device;
• database-backed inventory system;
• distributed file or task service;
• intrusion-detection demonstration using authorised data;
• edge-based image or sensor application.

Every project should state the problem, users, requirements, architecture, testing method, limitations and future improvement. Security and privacy should be designed from the beginning.