Loading

from first concept to confident practice

Machine Learning

Teach models from data, measure what they learn, and monitor what happens next.

Start learning 11 lessons · first lesson about 18 min · free · no account needed
About this track

Learn how machines learn from examples - and how to tell whether they really did. You’ll frame problems, prepare features without leaking the answer, and train linear and logistic regression with gradient descent you write yourself.

Then you’ll watch overfitting happen, choose models with cross-validation, read confusion matrices and ROC curves, grow decision trees and ensembles, cluster with k-means, build a small neural network, and monitor a deployed model for drift and uneven errors. Interactive widgets let you tune learning rates, raise polynomial degrees and step through k-means; every lesson ends with Python (and numpy) that runs in your browser, and the capstone builds a complete loan-default classifier.

Before you start
  • No prior machine learning experience needed.
  • Basic Python helps; numpy is introduced as you go.
  • Comfort with school algebra; the formulas are explained step by step.

Your path

0 of 11 lessons · 0%

Restoring your progress…

Loading progress…
  1. Unit 1 · 0/4 lessons

    ML foundations

    Frame problems, prepare features, and train linear and logistic regression with gradient descent. Badge: Data ready

    1. 0x001Understand machine learning problemsFrame a problem as classification, regression or clustering, and name its examples, features and target. 18 min 0/2 exercises solved 18 min 0/2 exercises solved
    2. 0x102Prepare data and featuresSplit data first, scale and encode features with training statistics only, and avoid leakage. 22 min 0/2 exercises solved 22 min 0/2 exercises solved
    3. 0x203Linear models and gradient descentFit a line by minimizing a loss, step by step, and see how the learning rate decides whether training works. 25 min 0/2 exercises solved 25 min 0/2 exercises solved
    4. 0x304Classify with logistic regressionTurn a linear score into a probability with the sigmoid, train with log loss, and pick a decision threshold. 24 min 0/2 exercises solved 24 min 0/2 exercises solved
  2. Unit 2 · 0/3 lessons

    Training and evaluation

    Fight overfitting, choose metrics, and build trees and ensembles. Badge: Model evaluator

    1. 0x405Train models and generalize to new dataSee overfitting happen, choose model complexity with validation data and cross-validation, and rein it in with regularization. 25 min 0/2 exercises solved 25 min 0/2 exercises solved
    2. 0x506Choose useful evaluation metricsRead a confusion matrix, compute precision, recall and F1, use ROC AUC, and pick regression metrics. 24 min 0/2 exercises solved 24 min 0/2 exercises solved
    3. 0x607Decision trees and ensemblesSplit data with Gini impurity, see why single trees overfit, and combine many trees into random forests and gradient boosting. 25 min 0/2 exercises solved 25 min 0/2 exercises solved
  3. Unit 3 · 0/3 lessons

    Patterns and deployment

    Cluster without labels, build a neural network, and monitor models in production. Badge: Responsible learner

    1. 0x708Find structure without labelsRun k-means by hand, choose k with the elbow method, and compress features with PCA. 24 min 0/2 exercises solved 24 min 0/2 exercises solved
    2. 0x809Neural networks from the ground upStack linear layers with nonlinear activations, run a forward pass, and see why a hidden layer solves XOR. 24 min 0/2 exercises solved 24 min 0/2 exercises solved
    3. 0x9010Deploy, monitor, and govern ML systemsDetect drift with the population stability index, check performance across subgroups, and plan responses before launch. 22 min 0/2 exercises solved 22 min 0/2 exercises solved
  4. Unit 4 · 0/1 lessons

    Capstone

    Build a complete loan-default classifier with numpy. Badge: ML practitioner

    1. 0xA011Capstone: build a loan-default classifierSplit, standardize, train logistic regression, and evaluate on a held-out test set - end to end with numpy. 35 min 0/2 exercises solved 35 min 0/2 exercises solved