Regression Explorer ŷ = w·x + b · mean squared error
Task
Model
loss: mean squared error
Learning rate
0.050
log scale · 0.001 – 10
Gradient descent
25/s
click the cost landscape to pick a start, or press Play
Data
click to add a point · click a point to remove it
Data space x · y
n = 0
Cost landscape J(w, b) — MSE
steps: 0J: —
Convergence
J
w
b
steps
grid min J