References
Notes {{word-count}}
Summary:
Key points:
What is Deep Learning?
Deep Learning is Machine Learning with multiple layers of learned representations.
In these layers of representations, each layer it performs a simple parameterized transformation (function) of the preceding layer.
As the layers go up, we have higher-level representations, which are more abstract, more invariant to n
This is a definition by Professor Sergey Levine at 25:25 in the second part of lecture 1.
The function that represents the transformation form input to internal representation to output is usually a deep Neural Network.
What makes Deep Learning work?
Big models with many layers.
Large datasets with many examples.
Enough compute to handle both.