Machine Learning
What machine learning actually is, taught from zero with no hand-waving — from vectors, probability and Bayes' theorem, through every major classical model, to neural networks and the real tools (NumPy, pandas, scikit-learn, PyTorch) used to build them.
Written byRohan Sharma
Foundations
3 lessons- 7 min read
What machine learning actually is
The one-sentence idea behind every ML system, explained with no jargon before we add any.
- 8 min read
The three flavors of learning
Supervised, unsupervised, and reinforcement learning — what separates them and which real problems belong to each.
- 9 min read
How a model actually learns
The guess, check, adjust, repeat loop that every ML model runs, shown first with a guessing game and then with real numbers.
Foundations Quiz
6 questions
Math Foundations
4 lessons- 9 min read
Vectors and the dot product
Why ML represents everything as lists of numbers, and the one operation (the dot product) that shows up in almost every model.
- 10 min read
Derivatives and gradients, without the fear
What a slope actually measures, why it tells you which way to move, and how gradient descent turns that into a training algorithm.
- 14 min read
Probability, conditional probability, and Bayes' theorem
What a probability actually is, how to combine and update them, and the theorem that powers an entire family of classic ML models.
- 10 min read
Descriptive statistics and feature scaling
Mean, median, variance, correlation, and why almost every model needs its input features rescaled before training.
Math Foundations Quiz
12 questions
Supervised Learning
6 lessons- 11 min read
Linear regression, from a straight line to real math
Building the simplest predictive model there is, and deriving exactly how it learns using the mean squared error and gradient descent.
- 10 min read
Logistic regression, predicting a category instead of a number
Reusing the linear regression machinery, adding the sigmoid function, and deriving the log loss that trains a classifier.
- 10 min read
Decision trees, learning by asking questions
A completely different kind of model with no weights or gradients at all — just a sequence of yes/no questions, chosen using entropy and information gain.
- 8 min read
K-nearest neighbors, predicting by asking your neighbors
A model with no training step at all, no weights, and no gradients — just the distance formula and a vote, and why that makes it both simple and slow.
- 8 min read
Building a real Naive Bayes spam classifier
Turning the Bayes' theorem math from the foundations module into a complete, working text classifier, end to end.
- 10 min read
Support vector machines, the widest possible margin
A classifier built around a genuinely different goal - not just separating two classes, but separating them with as much breathing room as possible.
Supervised Learning Quiz
14 questions
Training & Evaluation
4 lessons- 9 min read
Overfitting, underfitting, and regularization
Why a model that's perfect on training data can be useless in the real world, and the math that fixes it.
- 13 min read
Evaluation metrics, and why accuracy lies to you
Precision, recall, F1, ROC/AUC, and the confusion matrix, motivated by a classifier that looks great and is actually useless.
- 8 min read
The bias-variance trade-off
The single framework that explains why models underfit or overfit, and how it ties overfitting, regularization, and model complexity into one idea.
- 11 min read
Data preparation and hyperparameter tuning
What happens before training even starts - handling missing values, encoding categories - and how to systematically search for the best settings.
Training & Evaluation Quiz
13 questions
Unsupervised Learning & Ensembles
3 lessons- 9 min read
K-means clustering, finding groups nobody labeled
How a computer sorts data into groups with no answer key at all, using nothing but the distance formula from the math foundations module.
- 9 min read
PCA, squashing many features into a few that matter
How principal component analysis finds the handful of directions in your data that carry almost all the real information, and drops the rest.
- 10 min read
Ensembles, random forests, and boosting
Why combining many weak, imperfect models often beats any single strong one, using nothing more than averaging and voting.
Unsupervised Learning & Ensembles Quiz
8 questions
Neural Networks
4 lessons- 9 min read
The perceptron, a single artificial neuron
The tiny building block behind every neural network, built from nothing but the dot product and sigmoid you already know.
- 11 min read
Backpropagation, training a network with many layers
How the chain rule lets a single algorithm assign blame to every weight in a deep network, all the way back to the first layer.
- 10 min read
Training neural networks in practice
What actually changes when you go from plain gradient descent to training a real network - better optimizers, batching, dropout, and batch normalization.
- 8 min read
The deep learning landscape, and where to go from here
How CNNs, RNNs, and transformers are all the same perceptron-and-backpropagation foundation, wearing different architectural clothes.
Neural Networks Quiz
12 questions
Tools & Practice
6 lessons- 10 min read
NumPy and pandas, the two libraries under everything
The real tools used to do vector math and load real datasets, instead of the plain-Python loops used throughout this course.
- 8 min read
Visualizing data and models with Matplotlib
Turning every diagram drawn in text throughout this course into real plots you generate from your own data and models.
- 10 min read
scikit-learn, every classical model in a few lines
How linear regression, logistic regression, trees, forests, SVMs, k-means, and PCA look in the library actually used to build them in practice.
- 11 min read
PyTorch, from your hand-written network to a real framework
Mapping every piece of the perceptron and backpropagation lessons directly onto PyTorch's tensors, autograd, and training loop.
- 12 min read
An end-to-end project walkthrough
Every lesson in this course, applied in order, to one real dataset from raw CSV to a trained, honestly-evaluated model.
- 6 min read
Cheat sheet: NumPy, pandas, Matplotlib, scikit-learn, PyTorch
A single quick-reference page for every tool covered in this module - the commands you'll actually reach for, without the explanations.
Tools & Practice Quiz
8 questions
Assessments
8 quizzesFoundations Quiz
Foundations · 6 questions
Math Foundations Quiz
Math Foundations · 12 questions
Supervised Learning Quiz
Supervised Learning · 14 questions
Training & Evaluation Quiz
Training & Evaluation · 13 questions
Unsupervised Learning & Ensembles Quiz
Unsupervised Learning & Ensembles · 8 questions
Neural Networks Quiz
Neural Networks · 12 questions
Tools & Practice Quiz
Tools & Practice · 8 questions
Machine Learning Final Assessment
Final assessment · 20 questions