Mark Mote

Infinitely differentiable, very smooth functions

k(x, x') = \sigma^2 \exp\left(-\frac{(x-x')^2}{2\ell^2}\right)

Kernel shape k(0, x):

Covariance matrix K(x, x′):

Length Scale: 0.50

How quickly correlations decay with distance

Signal Variance: 1.00

Overall variance of the function

Noise Variance: 0.010

Expected noise in observations

Show posterior meanShow uncertainty bandShow sample functionsShow data points

Click on the plot to add data points

Drag points to move them

Posterior Mean

95% Confidence

Observations

Click anywhere to add data points