A hidden smooth function is drawn from a Gaussian process with a squared-exponential kernel. Click on the plot to observe a point; the posterior mean and its ±2σ confidence band are recomputed from the kernel. Watch the band tighten around the data and relax between them. The length-scale ℓ controls how smooth the model believes the function to be.
Shaded region = ±2σ* (about 95% of the posterior mass). Dots = observations. Dashed grey = the true (hidden) function.