Interactive health ML

See how models really work.

Explore complex machine-learning ideas through interactive explanations, visual experiments, and health-focused examples.

Explore the roadmap Test your knowledge

40+ visual lessons73 core and 32 advanced quiz questionsHealth ML examples

Biomedical ML lifecycle Evidence before deployment

1
Define the cohort and outcome EHR, imaging, genomics or sensors · eligibility · time zero
2
Preprocess without leakage Impute, encode and scale inside each training partition
3
Develop and tune the model Compare candidates · regularise · use nested resampling
4
Validate performance and transportability Internal + external validation · calibration · subgroups
5
Assess utility, impact and monitoring Thresholds · net benefit · prospective impact · model drift

Explore the library

Learn by concept.

Choose a topic and move at your own pace. Each lesson pairs concise theory with an interactive or visual explanation.

🧭 Learning Navigation 💡 Start here! Use the roadmap to plan your learning journey, then test your knowledge with the quiz.

Interactive Learning Roadmap

Navigate your ML journey with an interactive mindmap organized by topics, difficulty levels, and learning paths

ML Knowledge Quiz

73 original questions, plus an advanced set on neural networks and transformers. Filter by topic or practice one question at a time

Medical statistics (BA2)

Logistische regressie lab

Dutch SPSS-style frequencies, crosstabs, and logistic regression in the browser for Year 2 medicine practicals

🤖 Core Machine Learning Concepts

ML Algorithms Overview

Introduction to machine learning algorithms and their applications

ML Algorithms Comprehensive Guide

Complete guide to machine learning algorithms with examples and practical applications

ML Algorithms Mathematical Guide

Mathematical foundations and detailed formulations of machine learning algorithms

Geometric Interpretation of ML

Visual understanding of machine learning algorithms

ML Algorithm Animations

Step-by-step visual lab for linear regression, clustering, trees, neural networks, random forests, and logistic regression

Cost Functions & Loss Visualization

Understanding optimization through cost functions

Optimization Algorithms

Different approaches to model optimization

ML Algorithm Selection Flowchart

Interactive guide to choosing the right ML algorithm

📊 Data & Mathematical Foundations

Data as a Matrix

Understanding how data is structured as matrices

Non-Invertible Matrices

When matrices cannot be inverted and why it matters

🧪 Data Preparation & Pipelines

Missing Data, Imputation & Safe ML Pipelines

Practical guide to preprocessing health datasets without leakage, from missing values to train-only pipelines

Label Quality & Outcome Definition in EHR Data

Student-friendly guide to defining outcomes, spotting noisy proxy labels, and avoiding label leakage in health ML

Privacy, Governance & Data Sharing in Health ML

Student-friendly guide to minimum necessary data, sharing routes, approvals, audit trails, and privacy-aware collaboration design

📈 Dimensionality Reduction

Comprehensive guide to Principal Component Analysis with examples

Medical Autoencoder Explainer

PCA vs UMAP of BloodMNIST latents: click cells, interpolate, and colour by type, error, or simulated site

📊 Model Evaluation & Validation

Train, Validation, Test Splits & Data Leakage

Foundational guide to safe data splitting, leakage prevention, and realistic evaluation in health ML

External Validation & Transportability

Healthcare guide to testing whether models generalize across hospitals, populations, and time periods

Deployment, Monitoring & Model Drift

Health ML guide to post-deployment surveillance, drift detection, alert fatigue, rollback, and recalibration decisions

Model Updating, Recalibration & Lifecycle

How to decide whether a health ML model needs investigation, recalibration, refitting, or retirement after deployment

Clinical Workflow Integration & Human-AI Decision Support

Healthcare guide to where predictions enter care pathways, who acts on them, alert burden, oversight, and safe workflow design

Prospective Validation, Impact Studies & Randomized Evaluation

Health ML guide to silent trials, pilot rollouts, pragmatic comparisons, and how to test whether a model changes care safely

Reporting Standards, Reproducibility & Model Cards

Capstone guide to transparent health ML write-ups, reproducible analysis logic, model cards, and honest deployment claims

Bias-Variance Tradeoff

Understanding model complexity and generalization

Bias-Variance Applications

Applied examples of bias-variance tradeoff

ROC Curves & Performance Metrics

Evaluating classification model performance

Precision, Recall & Class Imbalance

Health ML guide to rare outcomes, alert quality, prevalence shift, and precision-recall tradeoffs

Calibration, Thresholds & Clinical Decisions

Student-friendly guide to probability reliability, threshold selection, and action tradeoffs in health ML

Uncertainty & Prediction Sets Lab

After a probability, a prediction set can still say Unsure instead of forcing yes or no. The 90% target can fail in a new hospital

Survival Analysis & Competing Risks

Health ML guide to time-to-event outcomes, censoring, survival curves, and competing clinical events

Model Interpretability & Explainable AI

Healthcare-focused guide to global and local explanations, feature effects, and safe model reporting

ML Evaluation Metrics Table

Comprehensive overview of machine learning evaluation metrics

🔄 Cross-Validation Techniques

K-Fold Cross-Validation

Interactive visualization of k-fold validation

Nested Cross-Validation

Advanced validation for hyperparameter tuning

Training Process Visualization

Visual guide to the training and validation process

⚙️ Hyperparameter Optimization

Hyperparameter Tuning Guide

Comprehensive guide to optimizing model parameters

🎯 Ensemble Learning

Ensemble Methods Interactive

Interactive exploration of ensemble learning techniques

🧠 Neural Networks

NN vs Classical ML Training

Training approaches comparison between neural networks and classical ML

Neural Network Training Slides

Comprehensive guide to neural network training process

Parameter Growth in Neural Networks

Understanding how parameters scale with network size

Neural Network Issues & Challenges

Common problems and solutions in neural network training

Backpropagation Learning Process

Interactive visualization of how neural networks learn through backpropagation

Medical CNN Explainer

Interactive Tiny-VGG on blood-cell images: click feature maps to watch convolution, ReLU, pooling, and softmax

Dense EHR Explainer

Simulated tabular patients: compare logistic regression with a tiny MLP on interactions, thresholds, and U-shaped risk

Medical Autoencoder Explainer

Interactive BloodMNIST autoencoder: reconstruct cells, then play with PCA, UMAP, and a 32-D latent space — including a simulated site effect

Neural Network Architecture Lab

Edit one signal and trace dense, convolutional, recurrent, and attention information flow

Multimodal Fusion Lab

Train six models, explore 2D decision maps and 3D score surfaces, trace fusion architectures, and test noisy or mismatched modalities

Clinical Transformer Explainer

Tiny GPT on simulated notes: a left-to-right architecture flow, GPT-2 BPE playground, and next-word sampling — with healthcare caveats

Autoencoder Tutorial

Comprehensive guide to autoencoders with interactive examples and healthcare applications

Autoencoder Animations

Interactive visualizations of different autoencoder architectures with mathematical details and training guidelines

Biological vs Artificial Neural Networks

Interactive comparison between biological and artificial neurons with live calculator and activation function exploration

⚖️ Bias & Fairness in ML

Understanding Bias in ML

Comprehensive guide to bias types, detection, and prevention in machine learning

Causal Inference & Confounding

Healthcare-focused guide to separating association, prediction, and causal effects with DAGs and confounding examples

🧠 Advanced Topics

Federated Learning

Privacy-preserving collaborative AI training with interactive demonstrations

🆘 Help & Support

🚀 Getting Started

Welcome to the Interactive ML Learning Platform! Here's how to make the most of your learning experience:

Start with the Roadmap: Use the interactive roadmap to plan your learning journey
Take the Quiz: Test your current knowledge to identify areas for improvement
Follow Learning Paths: Topics are organized by difficulty level (Beginner → Intermediate → Advanced)
Interactive Elements: Click, hover, and explore to engage with the content

📚 Navigation Tips

Topic Categories: Content is organized into logical categories for easy browsing
Mobile Friendly: All content is optimized for mobile devices
Search Function: Use the roadmap search to find specific topics quickly

🎯 Learning Recommendations

💡 Pro Tip: Start with "Core ML Concepts" if you're new to machine learning, or jump to "Advanced Topics" if you have ML experience and want to explore deeper concepts.

Beginners: Start with Mathematical Foundations → Core ML Concepts → Data Preparation & Pipelines → Data Splitting & Leakage → Model Evaluation
Intermediate: Focus on Model Evaluation → Calibration & Thresholds → Precision, Recall & Imbalance → Cross-Validation
Advanced: Explore Neural Networks → Advanced Topics → Specialized Applications

🔧 Technical Support

If you encounter any issues:

Browser Requirements: Use a modern browser (Chrome, Firefox, Safari, Edge) with JavaScript enabled
Mobile Issues: Try rotating your device or using landscape mode for better viewing
Interactive Elements: Some animations may take a moment to load on slower connections
Content Updates: Check back regularly as new content is being added

📞 Contact & Feedback

For questions, suggestions, or to report issues:

Platform Creator: Prof. Gennady Roshchupkin
Platform: Interactive ML Learning Platform
Support: This is an educational resource - please be patient with any technical issues

🎓 Learning Goal: This platform is designed to make complex ML concepts accessible through interactive visualizations. Take your time, experiment with the tools, and don't hesitate to revisit topics as your understanding grows.