Artificial Neural Network

Supervised Training Visualizer
Iteration 0 / 40
Training Example 0 / 5
Current Loss
Current Phase Idle
0% Learning
activation = sigmoid( Σ (input × weight) + bias ) Σ means "add all the weighted inputs together."

What is happening now?

1Forward pass
2Compare
3Backpropagation
4Update

Current Example

Study Hours
Sleep Hours
Prediction
vs
Actual

Inside one learning step ?

Pause training, then click any hidden or output node in the diagram.

No node selected yet.

Before → After Update

Run a training step to see a weight and bias change.

Loss Before → After

Run a training step to compare loss before and after.

Try the Trained Network

This does not change any weights or biases — it only asks the network to make an estimate.

The ANN does not know the exact answer. It learns patterns from the examples and uses those patterns to make an estimate. With only five examples, predictions are rough estimates, not guaranteed facts.

Training summary — in simple words

Once you start training, this section will explain in plain language what the network learned and how.