SciML Notebooks

Interactive notebooks for teaching Scientific Machine Learning (ES98E and PX914) in the HetSys CDT and Predictive Modelling and Scientific Computing MSc, developed by James Kermode.

Students on the module: lectures and workshops run on the SciML notebook hub. Run a lecture, or edit and submit your own copy of a workshop, in the browser.

Open the SciML hub →

University of Warwick SSO login. The hub is offline at the moment: LIVE demos are unavailable; WASM and molab notebooks still work.

Demonstrations

Probability

Probability DistributionsWASMmolab

Regression

Gaussian processes

Kernel ExplorerWASMmolab
GP RegressionWASMmolab
GP with Derivative Observationsmolab
Bayesian Optimisationmolab

Neural networks

MLP Regressionmolab
BNN Regression (Variational Inference)molab
Neural ODE Solvermolab

Unsupervised learning

Uncertainty quantification

Uncertainty in Models and Data: Reaction KineticsDEMOmolab

WASM notebooks run in your browser (no login required). LIVE notebooks require University of Warwick SSO. DEMO presentations are public (no login required).

molab links open notebooks in marimo's free cloud environment — no login or installation required. Useful for external collaborators and students without a Warwick account.