Neural Network
Training Weights
Epoch
0
| Loss
1.000
| LR
0.010
Network Architecture
Hidden Layers
3
Neurons per Layer
6
Input Features
4
Output Classes
3
Training
Learning Rate
0.010
Training Speed
1x
Noise / Regularization
0.10
Train
Pause
Reset
Display
Weight Thickness
1.0
Neuron Scale
1.0
Positive weight
Negative weight
Near-zero weight
Low error neuron
Medium error
High error neuron
Line opacity = weight magnitude
W
A
S
D
move
Space
/
Shift
up/down Right-click + drag to look Scroll to zoom