Introduction
What is the purpose of this lecture?
In the previous lectures, you have seen how to use different Machine Learning libraries for supervised problems. Deep Learning methods correspond to a subclass of Machine Learning techniques. This lecture provides an introduction to a subclass of Machine Learning called Deep Learning. And we will show how to use 2 popular Deep Learning libraries (PyTorch and TensorFlow) on an simple image-recognition task.
Plan
First, we will start by providing a brief introduction to Deep Learning. In particular, we will explain the main principles of Deep Learning. And we will show how to apply these principles on an easy image-recognition task.
Then, we will see how PyTorch and TensorFlow can be applied to that same task in practice. Also, we will highlight the similarities and differences between those Deep Learning libraries.
Finally, we will show which components of a Deep Learning technique can be modified. And we will provide various examples of other Deep Learning problems.
[All credit goes to Luca Grillotti for preparing all these materials!]