Description

Course Description

This 6-month certification program provides a comprehensive and hands-on introduction to Machine Learning (ML). Students will learn the fundamental concepts, algorithms, and techniques behind modern ML systems while also gaining practical experience through coding labs,
mini-projects, and a final guided capstone project.

The course blends theory, mathematics, and programming practice to build a strong foundation. By the end of the program, learners will be able to design, implement, evaluate, and deploy ML solutions across real-world domains such as healthcare, finance, natural language processing (NLP), and computer vision, with an awareness of AI ethics, bias, and fairness.

Intended Audience

• Undergraduate students (3rd year and above) in Computer Science, IT, Electronics, Data Science, or related fields
• Postgraduate students seeking applied ML skills
• Early-career professionals or enthusiasts aiming to switch to AI/ML career tracks
• Learners with basic programming knowledge who want structured training with practice

Prerequisite Knowledge

To succeed in this course, learners should have:

• Mathematics: High school level linear algebra, probability, and calculus basics (refresher provided)
• Programming: Basic knowledge of Python (variables, loops, functions, lists, dictionaries)
• Statistics: Basic descriptive statistics and probability distributions
• Enthusiasm for solving problems with data and exploring AI

Course Highlights

• Balanced Curriculum: Covers ML theory, algorithms, practical applications, and ethics
• Hands-On Learning: Weekly coding labs, assignments, quizzes, and mini-projects
• Capstone Project: Guided 2-week project with real-world datasets from healthcare, finance, NLP, or vision
• Industry-Relevant Skills: Feature engineering, model building, deep learning, deployment, and explainability
• Tools & Libraries: Python, NumPy, Pandas, Matplotlib, Scikit-Learn, TensorFlow/Keras, PyTorch (intro)
• Ethics & Responsible AI: Learn bias detection, fairness evaluation, and explainable AI techniques
• Flexible Learning: Combination of online conceptual classes and offline practical sessions
• Portfolio Ready: Students graduate with projects, a deployed ML app, and a capstone project to showcase

Module 1: Foundational Skills and ML Introduction (16 hrs)

Topics

• Introduction to ML applications
• Supervised vs. unsupervised learning paradigms
• ML pipeline
• Ethics in AI
• Python refresher (NumPy, Pandas, Matplotlib, Seaborn).
• Git basics
• Linear Algebra (vectors, matrices, dot product)
• Probability & Stats (distributions, Bayes’ Theorem)
• Calculus (derivatives, gradients)
• Optimization (GD, SGD, Adam, learning rate)

Learning Outcomes

• Understand the fundamental concepts of machine learning and its applications.
• Be proficient in using Python libraries for data manipulation and visualization.
• Grasp the core mathematical concepts that underpin ML algorithms.
• Set up a professional development environment with version control.

Labs / Hands-on

• Python and Jupyter notebook setup.
• Data exploration and visualization using a real-world dataset.
• Implementing basic linear algebra and calculus operations in Python

Assessment

• Quiz 1: Covers introductory concepts, Python basics, and Git commands.
• Lab Assignment 1: Analyze and visualize a dataset, submitting the notebook via Git.
• Mini-Project: A small project involving data cleaning and basic visualization.

Module 2: Supervised Learning & Model Evaluation (16 hrs)

Topics

• Data cleaning, missing values, encoding, normalization
• train-test split
• Linear and Polynomial Regression
• Regularization (Ridge, Lasso)
• Regression metrics (RMSE, R2)
• Logistic Regression
• k-NN, Decision Trees, Naive Bayes
• Classification metrics (accuracy, precision, recall, F1, ROC)
• Bias-variance tradeoff
• Cross-validation and hyperparameter tuning.

Learning Outcomes

• Prepare data for supervised learning models.
• Build and evaluate both regression and classification models.
• Understand and apply various evaluation metrics.
• Tune hyperparameters to improve model performance.

Labs / Hands-on

• Data preprocessing and feature engineering.
• Implementing and comparing different regression models.
• Building and evaluating various classification models.
• Using cross-validation and hyperparameter tuning on a chosen model.

Assessment

• Quiz 2: Covers data preprocessing, regression, and classification theory.
• Lab Assignment 2: Implement a complete regression pipeline, including preprocessing and model evaluation.

Module 3: Advanced Supervised & Unsupervised Techniques (16 hrs)

Topics

• Ensemble methods (Bagging, Boosting, Random Forest, AdaBoost, Gradient Boosting, XGBoost)
• Unsupervised learning (k-Means, Hierarchical, DBSCAN)
• Anomaly detection
• Dimensionality reduction (PCA, LDA, t-SNE)

Learning Outcomes

• Utilize powerful ensemble methods to build more accurate models.
• Apply unsupervised learning techniques for clustering and anomaly detection.
• Reduce data dimensionality for improved model performance and visualization.

Labs / Hands-on

• Implementing and comparing different ensemble methods on a classification task.
• Applying clustering algorithms to an unlabeled dataset.
• Using PCA for dimensionality reduction and visualizing the results.

Assessment

• Quiz 3: Covers ensemble methods, clustering, and dimensionality reduction.
• Mini-Project: Perform an unsupervised learning task (e.g., customer segmentation) on a dataset.

Module 4: Neural Networks and Deep Learning (16 hrs)

Topics

• Perceptrons
• Feedforward NNs
• Backpropagation
• activation functions
• loss functions
• Regularization (dropout, batch norm)
• CNN basics (convolution, pooling)
• Architectures (LeNet, AlexNet, ResNet)
• Transfer learning
• Intro to RNNs/LSTMs

Learning Outcomes

• Understand the fundamentals of deep learning and neural networks.
• Build and train a simple neural network from scratch.
• Apply CNNs for image classification and utilize transfer learning.
• Grasp the basics of processing sequential data with RNNs/LSTMs.

Labs / Hands-on

• Building a simple feedforward neural network.
• Implementing a CNN for an image classification task.
• Using a pre-trained model with transfer learning.

Assessment

• Quiz 4: Covers deep learning concepts, CNNs, and RNNs.
• Lab Assignment 3: Build a complete image classification model using a CNN and transfer learning.

Module 5: Specialized ML and Deployment (16 hrs)

Topics

• Intro to NLP (Bag-of-Words, TF-IDF)
• Word embeddings (Word2Vec, GloVe)
• Transformers
• Time series (ARIMA, LSTMs)
• SVM
• Autoencoders
• Recommender Systems
• Reinforcement Learning
• Flask/Streamlit deployment
• Docker basics
• MLOps concepts (CI/CD, model monitoring, reproducibility).

Learning Outcomes

• Process and analyze text data using NLP techniques.
• Develop a basic time series forecasting model.
• Understand specialized models like SVMs and Recommender Systems.
• Deploy a machine learning model as a web application.

Labs / Hands-on

•  NLP lab on text processing and sentiment analysis.
• Building a time series forecasting model.
• Deploying a trained model using Flask or Streamlit.

Assessment

•  Quiz 5: Covers NLP, time series, and deployment concepts.
• Mini-Project: Build a recommender system or a time series forecasting model and present the results.

Module 6: Case Studies and Capstone Project (16 hrs)

Topics

• Case studies across domains (CV, NLP, healthcare, finance)
• Project scoping and proposal writing
• Dataset exploration and planning
• Implementation, evaluation, and presentation

Learning Outcomes

• Integrate all learned skills to solve a real-world problem.
• Manage a complete ML project lifecycle from start to finish.
• Effectively communicate project findings.
• Develop a porfolio-worthy project.

Labs / Hands-on

• This entire module is a hands-on, project-based learning experience. Students work on a single, comprehensive capstone project.
  Assessment
• Project Proposal: A written document outlining the problem, dataset, and planned approach.
• Final Project: The completed ML project, including code and a polished report.
• Final Presentation: A presentation of the project, including the problem, methodology, and results.

Sample Project Ideas

• Predict disease risk or patient outcomes from structured or unstructured medical data.
• Sentiment analysis, fake news detection, or topic modeling on social media data.
• Build an image classifier or object detector for a domain-specific application.
• Stock price/cryptocurrency forecasting, fraud detection, or credit risk modeling.
• Predict traffic congestion, pollution, or energy usage.

Assessment

•  Final Project Evaluation (40%)
• Viva & Presentation (10%)

 

Course Assessment Scheme (6 Months)

• Quizzes (5 × 4%) → 20%
• Assignments & Lab Reports → 15%
• Mini-projects → 15%
• Final Project (report + code + presentation + viva) → 50%