In the past few years, deep learning (DL) has emerged as a powerful AI method that has found applications in areas such as object recognition, image segmentation, video analysis, and natural language processing. This course will discuss where and how deep learning is used and how to get people to use it. The approach will minimize the math and concentrate instead on the underlying ideas and principles. We will concentrate on Tensorflow/Keras as the underlying computational platforms and use Python to create the DL codes. Much of the course will be driven by a number of hands-on exercises that will help you build simple networks in Keras/Tensorflow. These exercises will cover analysis of tabular data, images, videos, and text. At the end of the course you will have a basic understanding of DL, Tensorflow/Keras as a DL platform, and example applications.
Deep Learning appears to be a magical algorithm that can solve difficult problems in a variety of domains. The coming together of a number of trends underlie the success of DL. In this course, you will get a chance to look under the hood of the DL hype and see the underlying architectures, and you will be exposed to a set of tools that you can use to create your own deep learning models.
Day 1:
- Introduction to Neural Networks (NNs) and Deep learning (DL)
- Training and Evaluation of Deep Neural Networks (DNNs)
- Tensors, Tensorflow and Keras
- Convolutional Neural Networks (CNNs)
- Recurrent Neural Networks (RNNs)
Day 2:
- Autoencoders (AEs)
- Transfer Learning
- Transformers
- Diffusion Models
This course will assume an understanding of statistics and calculus at the undergraduate level and programming experience with Python to get full benefits from the class.
Make sure to complete the follow pre-class checklist:
- Create a Google Account if you don't have one.
- If you are not familar with Google Colab, complete these tasks:
- Go over this basic tutorial. Some of the interface has changed since the tutorial was created but it should still be easy to follow.
- You are welcome to follow the rest of the links but they are not required.
- Go over the content of this interactive Colab notebook from the NVIDIA Deep Learning Institute introducing some basic concepts.