Skip to content
Intura
enid
Courses / Machine Learning Fundamentals

Course

Machine Learning Fundamentals

The habits that keep a model honest and the classic algorithms, each with a runnable scikit-learn example.

Lessons
30
Reading time
6 h 9 min
Level
Beginner
Taught in
English
Format
Articles with runnable code

Course

What you will learn

Thirty article lessons. Foundations: splitting a dataset, data leakage and the bias–variance trade-off. Features: scaling, transforming and engineering them. Evaluation: classification and regression metrics and imbalanced data. Algorithms: linear and logistic regression, k-nearest neighbours, support vector machines, decision trees, random forests, ensembles and gradient boosting. Then hyperparameter tuning, regularisation, feature selection, feature importance, SHAP and partial dependence plots, clustering, dimensionality reduction, anomaly detection and a pipeline end to end. Every lesson ends with code you can run and the output it prints.

  • Machine learning foundations
  • Preparing features
  • Measuring performance
  • Linear and distance-based models
  • Trees and ensembles
  • Tuning and regularisation
  • Feature selection
  • Model interpretation
  • Unsupervised learning
  • Putting it together

Before you start

Python you can read and run, and NumPy-level comfort with arrays — the Python Programming for Data course covers both. No maths beyond school algebra is assumed; every formula is explained in words.

Syllabus

Syllabus

Every lesson explains one idea, then runs it in scikit-learn and shows the output. The first lesson is open to everyone.

01 · Machine learning foundations

  1. 01Dataset splittingTrain, validation and test sets, k-fold cross-validation, and splits that respect classes, groups and time.Free preview12 min
  2. 02Data leakageHow information from the test set or the future sneaks into training, inflates your score, and how pipelines stop it.10 min
  3. 03The bias–variance trade-offWhy models underfit or overfit, how to tell from learning and validation curves, and what moves each kind of error.12 min

02 · Preparing features

  1. 04Feature scalingPut features on comparable scales for distance- and gradient-based models: standard, min-max and robust scaling, fitted on train only.11 min
  2. 05Feature transformationReshape skewed numbers, encode categories, bin, impute and expand features, each on the right columns with ColumnTransformer.11 min
  3. 06Feature engineeringBuild new features from domain knowledge — ratios, aggregates, dates and cycles — without leaking the answer.12 min

03 · Measuring performance

  1. 07Classification metricsAccuracy and why it lies, the confusion matrix, precision, recall and F1, ROC and precision–recall curves, and log loss.14 min
  2. 08Regression metricsMAE, RMSE, R² and MAPE: what each one punishes, how to read residual plots, and which to report for which decision.15 min
  3. 09Imbalanced dataWhen one class is rare: stratified splits, class weights, resampling, threshold moving and metrics that do not hide the minority.14 min

04 · Linear and distance-based models

  1. 10Linear regressionFit a straight line to predict a number: the squared-error loss, how the weights are found, regularisation and reading the coefficients.10 min
  2. 11Logistic regressionTurn a linear score into a probability with the sigmoid, train it with log loss, and choose the threshold that suits the decision.10 min
  3. 12k-nearest neighboursPredict from the closest examples: distance metrics, choosing k, why scaling matters and where the curse of dimensionality bites.11 min
  4. 13Support vector machinesFind the widest margin between classes, trade it against mistakes with C, and bend the boundary with kernels.12 min

05 · Trees and ensembles

  1. 14Decision treesSplit the data with yes/no questions: Gini impurity, entropy, reading a fitted tree and stopping it before it memorises.10 min
  2. 15Random forestsAverage many decorrelated trees: bootstrap samples, random feature subsets and the out-of-bag score.10 min
  3. 16Ensemble learningCombine models so their errors cancel: bagging, boosting, voting and stacking, and when each one helps.10 min
  4. 17Gradient boostingAdd small trees that each fix the last one’s residuals: learning rate, number of trees, early stopping and HistGradientBoosting.10 min

06 · Tuning and regularisation

  1. 18Hyperparameter tuningSearch settings with grid and random search on cross-validation, use log-scale ranges, and get an honest estimate with nested CV.11 min
  2. 19Regularisation: ridge, lasso and elastic netPenalise large weights to stop a linear model overfitting: L2 shrinks, L1 zeroes, elastic net mixes both, and alpha sets the strength.14 min
  3. 20Regularisation beyond linear modelsThe same idea in other models: tree depth and leaf limits, learning rate and subsampling in boosting, and early stopping.14 min

07 · Feature selection

  1. 21Feature selection: filter methodsDrop features before modelling with variance thresholds, correlation, mutual information and univariate tests, inside a pipeline.14 min
  2. 22Feature selection: wrapper and embedded methodsLet the model choose: recursive feature elimination, sequential selection, L1 and tree-based SelectFromModel, and their costs.14 min

08 · Model interpretation

  1. 23Feature importanceFind what a model relies on: coefficients, impurity and permutation importance, partial dependence and SHAP, and where each misleads.13 min
  2. 24SHAP valuesSplit one prediction into each feature’s contribution with Shapley values: waterfall, beeswarm and dependence plots, and their limits.15 min
  3. 25Partial dependence and ICE plotsSee how the prediction moves as one feature changes, on average and row by row, and spot interactions and correlated-feature traps.15 min

09 · Unsupervised learning

  1. 26k-means clusteringGroup rows with no labels: centroids and inertia, choosing k with the elbow and silhouette, and why scaling and initialisation matter.12 min
  2. 27Hierarchical clustering and DBSCANClusters of any shape: dendrograms and linkage, density-based clustering that finds noise, and choosing between them.12 min
  3. 28Dimensionality reduction: PCA, t-SNE and UMAPCompress many columns into a few: principal components and explained variance, and non-linear maps for plotting high-dimensional data.13 min
  4. 29Anomaly detectionFind the rows that do not fit: z-scores, isolation forest, local outlier factor and one-class SVM, and how to judge them without labels.14 min

10 · Putting it together

  1. 30Pipelines, end to endOne Pipeline from raw table to tuned model: preprocessing, selection, tuning, evaluation, interpretation and saving it for reuse.14 min

Feedback

Help us improve this course

Suggest a lesson, point out something unclear, or report a mistake in the code. A person reads every message and replies by email.

Start with