Deep Learning for Engineers
Interest in deep learning based applications has grown exponentially over the past few years, due to its surprising effectiveness. While deep learning may resemble magic, it is simply a collection of algorithms capable of finding useful representations of data, at multiple levels of abstraction.
Q: Why did the deep net cross the road?— Sarah Schwettmann (@cogconfluence) January 17, 2016
A: We don't know. But look, it did it really well…#deeplearning
This course will demystify the fundamentals of contemporary deep learning techniques, covering just enough theory to ground hands-on application. After the course, students should be comfortable tackling their own deep learning projects, whether for their work or as side projects.
- Introduction to Deep Learning and Review of Relevant Math & Statistics
- Building a Neural Network from Scratch
- Image Classification with Convolutional Neural Networks (CNNs)
- Optimization Techniques
- Training Neural Networks: Learning Rates, Activation Functions and Dropout
- Common Architectures
- Recurrent Neural Networks (RNNs) and Long Short Term Memory (LSTM) Networks
- Generative Adversarial Nets (GANs)
Projects and exercises
This course involves a number of hands-on programming exercises, implementing many deep learning components from scratch, as well as utilizing existing tools once we have developed a thorough understanding of them. The course will be in Python, given its high readability and being common in the deep learning field.
Participants are expected to be confident programmers, with a recommended 2 or more years of professional experience or equivalent. All exercises will be in Python. No math background is assumed beyond high school level, although some familiarity with linear algebra is advantageous (you may enjoy watching some of Grant Sanderson’s wonderful Essence of Linear Algebra videos beforehand).
Schedule and price
Deep Learning for Engineers will next run twice per week for 4 weeks in January 2019. The total price is $1800.