Browse Code
Gaussian Process model for fitting deterministic simulator output. Establish efficient and reliable likelihood optimization through hybridized DIRECT-BFGS and multi-start BFGS algorithms. Programming Language: Matlab.
Features
- Returns prediction, Y, and prediction uncertainty estimate, MSE, for any number of points
- User is free to select training data. Suggest user scales simulator input to [0,1]^d.
- Efficient and reliable likelihood optimization.
- 4 Likelihood optimization routines: DIRECT-BFGS, DIRECT-IF, 0.5d multi-start BFGS, and 2d+1 multi-start BFGS
- Squared exponential correlation matrix, R.
- Addition of 'nugget' to R for improved stability of R^-1 and |R| computation.
- Lower bound on nugget to minimize over-smoothing.
- Iterative regularization method for improved accuracy when using a nugget.
Project Samples
Other Useful Business Software
Full-stack observability with actually useful AI | Grafana Cloud
Built on open standards like Prometheus and OpenTelemetry, Grafana Cloud includes Kubernetes Monitoring, Application Observability, Incident Response, plus the AI-powered Grafana Assistant. Get started with our generous free tier today.
Rate This Project
Login To Rate This Project
User Reviews
Be the first to post a review of GPMfit!
