A.:Deep Learning differs from other approaches in Artificial Intelligence in at least three respects.

First, it learns from data, as opposed to methods that require knowledge to be hardwired into the system.

Second, it can learn from raw data, as opposed to methods that require pre-processing or feature extraction to proceed.

Finally, it employs a deep neural net architecture (many layers), as opposed of methods that use just a few layers of neurons.