Learn core concepts, theory, and intuitive visualizations.
Build and train real ML models with clinical markers.
Interactive demos on clinical AI ethics, regression, and model transparency.
Core concepts to understand before exploring the implementation pipeline.
The specific data points used by a model to find patterns and make predictions.
The specific value or category the model is trying to learn and predict.
The broad science of making machines mimic human intelligence and reasoning.
A subset of AI where machines "learn" patterns directly from data.
Training with a teacher: The model sees data and the correct answer (Label).
Finding paths in the dark: The model groups data without being told the answer.
Hiding part of the data to test if the model actually "knows" its stuff.
A three-way split to fine-tune the model "settings" before the final exam.
When a model is too simple to capture the complexity of the data.
When a model memorizes "noise" instead of learning the actual trend.
The trade-off between strict assumptions and sensitivity to data noise.
Rotating data so every piece is used for both training and testing.
Opening the "Black Box" to see exactly how and why AI made a decision.
Ensuring algorithms are fair, unbiased, and respectful of human dignity.
Combining many models together to improve overall accuracy and stability.
Complex networks inspired by the human brain for processing high-dimensional data.
Training AI across institutions without ever moving raw patient data.
The architecture that powers modern AI by focusing on relevant patterns.
Massive models trained on almost all human text to understand language.
A specific type of Transformer designed to generate coherent, human-like text.
The ability of AI to create entirely new content (text, images, synthetic data).
Visual aids to grasp complex statistical trade-offs in Machine Learning.
Watch how model complexity captures noise.
Bias (Accuracy) vs Variance (Precision).
Hands-on demonstrations of where clinical AI fails, misleads, and discriminates.
Predicting continuous health markers like SBP from BMI and Age.
Visualizing probability thresholds for binary risk (e.g., Diabetes).
Interactive logic flows for cancer triage and diagnostic rules.
The "Consensus Board" - aggregating multiple diverse tree models.
Mastering the trade-off between Sensitivity and Specificity.
Understanding "Missing a Diagnosis" vs "False Alarms".
Visualizing the danger of Over-diagnosis vs Under-diagnosis.
How rotating data ensures clinical reliability and robustness.
Why larger models sometimes perform better (Double Descent).
Train a multi-layer neural network epoch-by-epoch. Watch train vs validation loss curves and detect overfitting live.
Explore self-attention on a clinical note. Click any token to see which terms the model focuses on for diagnosis.
Train a Q-learning agent to optimise sepsis vasopressor dosing. Watch cumulative reward grow as the agent learns.
Discover hidden patient subgroups without labels. Watch centroids migrate and find the optimal K using the Elbow method.
Build a taxonomy of patient data. Drag the cut threshold to define clusters and explore different linkage strategies.
Simplify high-dimensional clinical data. Project many variables onto Principal Components while preserving variance.
Fundamentals for Researchers: From Raw Data to Predictive Models.
Continuous outcomes (e.g. SBP).
Binary categories (e.g. Disease Risk).
Robust ensemble classification.
Complex non-linear patterns.
Logical flows for binary splits.
Grouping unlabelled data by distance.
Reducing dimensions to find core variance.
Nested grouping by distance hierarchy.
Density-based grouping with noise handling.
Note: This is a simulation showing how real-world ML files are generated.