Course Intro & Setting Up Google Colab

AI vs ML vs Deep Learning vs LLMs

These four terms are nested inside each other, like circles:

Term	What it means	Example
Artificial Intelligence (AI)	Any technique that makes computers behave "smartly"	Chess engines, route planners
Machine Learning (ML)	A subset of AI where the computer learns rules from data instead of being given them	Spam filters, house price prediction
Deep Learning (DL)	A subset of ML using neural networks with many layers	Image recognition, speech-to-text
Large Language Models (LLMs)	Deep learning models built on the Transformer architecture, trained on huge amounts of text	ChatGPT, Claude, Gemini

Rules vs learning

The spam filter is the clearest example of the difference:

• Traditional programming: a human writes the rule — "if the email contains 'lottery', mark it spam."
• Machine learning: we give the computer thousands of labelled emails (spam / not spam) and it finds the patterns itself.

Key idea: In ML, we provide data + answers and the computer produces the rules.

The course roadmap

Phase	Focus	You will build
0	Python, NumPy, Pandas, plotting, math intuition	Clean data analysis scripts
1	EDA, scikit-learn, Random Forests	3 end-to-end projects
2	Neural networks, Keras, PyTorch, CNNs	MNIST / CIFAR models from scratch
3	RNNs, LSTMs	Text generator or time-series forecaster
4	Attention, Transformers, GPT-2	A tiny GPT from scratch
5	Fine-tuning, LoRA, embeddings	Your own fine-tuned LLM

How to get the most out of it

1. Follow the order. Transformers only make sense once you understand neural networks, which only make sense once you understand gradients.
2. Type the code yourself. Don't copy-paste. Fixing your own errors is where most learning happens.
3. 1–2 hours a day beats 10 hours once a week.

Google Colab

Google Colab is a free, browser-based Python environment from Google.

• Nothing to install: it runs on Google's servers
• Free: all you need is a Google account
• Free GPU access, which we'll need from Phase 2 onwards

Getting started

1. Go to colab.research.google.com
2. Sign in with your Google account
3. Click New notebook
4. Rename it by clicking Untitled0.ipynb at the top

Cells

A notebook is made of cells:

• Code cells run Python. Run with the ▶ button or Shift + Enter.
• Text cells hold notes written in Markdown (# Heading, **bold**, - bullet).

print("Hello, Machine Learning!")

The last line of a code cell is displayed automatically, even without print:

365 * 24   # hours in a year → 8760

Useful things to know

Task	How
Run a cell	Shift + Enter
Add a code cell	+ Code (or Ctrl+M B)
Add a text cell	+ Text
Enable a GPU	Runtime → Change runtime type → T4 GPU
Where notebooks are saved	Google Drive → Colab Notebooks folder
Edit a course notebook	Open it, then File → Save a copy in Drive

Heads-up: Colab disconnects after a period of inactivity. Your code is saved, but variables are lost. Just re-run the cells from the top (Runtime → Run all).

Recap

• AI ⊃ ML ⊃ Deep Learning ⊃ LLMs
• ML learns rules from data instead of being given rules
• The course has 6 phases, from Python basics to fine-tuning LLMs
• Google Colab gives you free Python + GPU in the browser
• Code cells run Python; text cells hold Markdown notes

# 2. Print your name
print("Hello, I am Sudha!")

# 3. Minutes in a week: 7 days × 24 hours × 60 minutes
7 * 24 * 60   # 10080

# 4. Study hours: 1.5 hours/day × 30 days × 6 months
1.5 * 30 * 6   # 270.0