Artificial Intelligence & Machine Learning

AI problem types, learning paradigms, and real-world applications

Distinguish rules, classical machine learning, deep learning, and generative AI

Frame a business problem as a measurable AI task

Python syntax, functions, collections, files, environments, and notebooks

NumPy arrays, vectorized operations, and reproducible experiments

Write clean, testable data and modelling code

Descriptive statistics, probability, distributions, and sampling

Vectors, matrices, similarity, gradients, and optimization intuition

Interpret uncertainty, correlation, and model evidence responsibly

Load, clean, join, transform, and validate datasets with pandas

Handle missing values, outliers, leakage, imbalance, and categorical data

Explore relationships and communicate findings visually

Regression, classification, baselines, and feature engineering

Train linear models, trees, ensembles, and nearest-neighbour models

Select suitable algorithms for a problem and dataset

Train, validation, and test splits; cross-validation and pipelines

Accuracy, precision, recall, F1, ROC-AUC, and regression metrics

Tune models and diagnose underfitting, overfitting, bias, and variance

Clustering, dimensionality reduction, anomaly detection, and segmentation

Evaluate groups without labelled answers

Build and explain a customer or operational segmentation project

Neural networks, activation functions, loss, backpropagation, and optimizers

Build and train models with a modern deep-learning framework

Apply regularization and track experiments

Image representation, preprocessing, convolutional networks, and transfer learning

Classification and object-detection concepts

Evaluate a practical image model and its failure modes

Text cleaning, tokenization, embeddings, classification, and sequence models

Transformer and large-language-model foundations

Build and evaluate a text analysis application

Prompting, structured output, retrieval-augmented generation, and tool use

Ground models in trusted sources and measure response quality

Add safety, privacy, cost, and latency controls

Package models behind APIs and integrate them into applications

Version data and models; monitor drift, quality, latency, and failures

Reproduce, document, and maintain an ML system

Fairness, explainability, privacy, security, and human oversight

Identify misuse, harmful bias, and high-risk deployment contexts

Create a model card and responsible-use plan

Define a problem, dataset, baseline, metrics, and acceptance criteria

Build, evaluate, document, and deploy an end-to-end solution

Present results, limitations, risks, and recommended next steps