Artificial Intelligence (AI) — learn everything (complete learning path)

125 topics across 7 chapters

Chapter 1

Foundations of AI

Core AI concepts & terminology

2 subtopics

Types of learning: supervised vs unsupervised vs reinforcement (intuition)

Overfitting vs generalization (intuition)

Problem framing, data splits, and metrics (basics)

2 subtopics

Train/validation/test splits + data leakage patterns

Choosing the right metric (accuracy, F1, AUC, RMSE, etc.)

Search, planning, and constraints (classic AI)

2 subtopics

Heuristics and A* search (high level)

Constraint satisfaction problems (CSPs) basics

Experimentation and communication

2 subtopics

Reproducibility basics: seeds, configs, environments

Writing a model report: assumptions, results, limitations

Chapter 2

Math & Statistics for AI

Linear algebra for ML

2 subtopics

Matrices, norms, dot products (what you actually use)

Eigenvalues/SVD intuition and why they matter

Calculus and gradients

2 subtopics

Derivatives and partial derivatives (practice-focused)

Chain rule and backprop math (core idea)

Probability for ML

2 subtopics

Conditional probability and Bayes’ rule

Expectation and variance (how to compute + interpret)

Statistics and hypothesis testing

2 subtopics

Confidence intervals (intuition + simple calculations)

Hypothesis tests and p-values (practical meaning)

Optimization basics for ML

2 subtopics

Gradient descent family (SGD, momentum) intuition

Regularization as an optimization tradeoff

Chapter 3

Programming & Data Skills for AI

Python for AI (practical)

2 subtopics

Clean Python for ML: functions, modules, notebooks vs scripts

Debugging and profiling basics

Data wrangling (pandas-style skills)

2 subtopics

Groupby/joins/reshape skills for datasets

Data cleaning: missing values, outliers, duplicates

SQL and relational data basics

2 subtopics

SQL essentials: SELECT, JOIN, GROUP BY (hands-on)

Indexes and query performance (intuition)

Software engineering + Git for ML code

2 subtopics

Git basics: branches, pull requests, code reviews

Testing + project structure for ML codebases

Compute environment basics (Linux, GPUs, containers)

2 subtopics

Linux shell basics (files, processes, permissions)

GPUs + environments: conda/venv, Docker (concepts)

Chapter 4

Classical Machine Learning

Preprocessing & feature engineering

2 subtopics

Scaling/normalization + categorical encoding

Imbalanced data strategies (weights, resampling)

Supervised learning algorithms (core set)

2 subtopics

Linear regression and logistic regression (what + when)

Decision trees, Random Forests, gradient boosting

Unsupervised learning (core set)

2 subtopics

k-means clustering + PCA (core ideas)

Anomaly detection basics

Evaluation, validation, and tuning

2 subtopics

Cross-validation and stratification

Hyperparameter search (grid/random) + early stopping

Interpretability & debugging models

2 subtopics

SHAP/LIME overview (and common pitfalls)

Learning curves, ablations, and error analysis

Reinforcement learning (intro)

2 subtopics

MDPs, value functions, and rewards (intuition)

Q-learning vs policy gradients (high level)

Chapter 5

Deep Learning & Generative AI

Neural network fundamentals

2 subtopics

MLPs and activations (capacity + intuition)

Losses and backprop (intuition, not heavy math)

Training deep networks (practical)

2 subtopics

Normalization and dropout (why they help)

Optimizers (Adam/SGD) and learning rate schedules

Computer vision with CNNs

2 subtopics

Convolution, pooling, receptive fields (core concepts)

Transfer learning for vision models

Transformers for language and sequences

2 subtopics

Self-attention (what it computes)

Masking + encoder/decoder patterns

Generative models (GANs, VAEs, diffusion)

2 subtopics

GANs vs VAEs (when they’re used)

Diffusion models (denoising and sampling intuition)

LLM application building

3 subtopics

Prompting patterns: zero/few-shot, tools, structured outputs

Embeddings + RAG basics (retrieval and grounding)

Fine-tuning vs LoRA (when to choose)

Chapter 6

MLOps & Deployment

Pipelines, datasets, and reproducibility

2 subtopics

Dataset versioning, lineage, and quality checks

Workflow orchestration basics (DAGs, retries, artifacts)

Model packaging and serving

2 subtopics

Batch vs online inference patterns

Serving APIs (REST/gRPC) and latency budgets

Monitoring, drift, and incident response

2 subtopics

Data drift vs concept drift (detection intuition)

Observability + rollback plans (production mindset)

CI/CD for ML systems

2 subtopics

Automated tests for data + models

Artifacts, model registry ideas, and reproducible builds

Security & privacy in deployed AI

2 subtopics

Threat modeling for ML (prompt injection, model theft)

PII handling, access control, encryption basics

Scaling, latency, and cost

2 subtopics

Caching and batching to reduce cost/latency

Cost estimation and capacity planning (basics)

Chapter 7

Responsible AI & Capstone Projects

Fairness and bias

2 subtopics

Where bias comes from in real pipelines

100

Fairness metrics and tradeoffs (high level)

101

Privacy and data ethics

2 subtopics

102

Consent, data minimization, retention policies

103

Differential privacy (intuition) and limitations

104

Safety, robustness, and adversarial issues

2 subtopics

105

Adversarial examples (intuition) and common defenses

106

Safe evaluation: stress tests and red-teaming basics

107

Governance and documentation

2 subtopics

108

Model cards and datasheets (documentation habits)

109

Human-in-the-loop design and escalation paths

110

Capstone projects (portfolio builders)

3 subtopics

111

Capstone 1: Tabular ML project end-to-end

112

Capstone 2: Build an LLM RAG app + evaluation harness

113

Capstone 3: Deploy + monitor a model (production-like)

114

Career and continuous learning

3 subtopics

115

Portfolio: write case studies and show impact

116

Interview prep: ML fundamentals + basic system design

117

Staying current: reading papers and tracking new methods