Calculus Gradient Optimizer Studio

Optimizer Parameters

Algorithm

Learning Rate (α): 0.050

Momentum (β): 0.90

Max Iterations: 100

Calculus & Local Approximation

Local Linear Approximation:
$L(x,y) \approx f(a,b) + \nabla f \cdot (x-a, y-b)^T$

$$\nabla f = [\partial f/\partial x, \partial f/\partial y]$$

Trajectory Telemetry

Current Loss

0.0000

Steps Taken

Position (X, Y)

-1.50, 1.50

Status

Ready

Telemetry Export