Solution

The correct answer is Backward from sink to source.

Key Points

• Back propagation is a supervised learning technique used for training artificial neural networks.
• It involves adjusting the weights of the network to minimize the error between the actual output and the expected output.
• The process works by propagating the error backward from the output layer (sink) to the input layer (source), thereby updating the weights in the opposite direction of the forward pass.

Additional Information

• Option 1: Forward from source to hidden nodes - This describes the forward pass of data through the network, not the backpropagation process.
• Option 3: Forward from source to sink - Again, this describes the forward pass of data from the input layer to the output layer, not the backpropagation process.
• Option 4: Backward from sink to hidden nodes - While partially correct, this option does not fully describe the backpropagation process as it stops at the hidden nodes and does not mention the input layer (source).