Haslam College of Business Deep Learning

Registration

After clicking the “Registration” tab, please select “Operations Excellence” as the Program Category in order to select this Deep Learning Course.

Learning format

This synchronous/live hands-on course, featuring direct access to the instructor, is designed for business analytics alumni and those with similar knowledge who wish to start using deep learning in their organizations. Virtual meetings occur twice per week during business hours, with each meeting lasting 75 minutes. This course also offers office hours, enabling students to obtain advice on their current deep learning company projects.

Bring your own project

Learners will be asked to complete an individual semester-long deep learning project and write a paper. This can be a project that a learner is currently working on or planning to start working on in their company. Open and one-on-one office hours will be offered on a weekly basis to talk through math, coding, and conceptual questions and issues.

Learning objectives

• TensorFlow 2 in Python
• Computing with massive datasets using recursive estimation
• Deep supervised learning
• Predictive applications in business

High-level overview of topics covered in this course 

Introduction

• What are deep neural networks?
• When should I use deep neural networks?
• How can I look inside the deep neural network black box?
• Why should I use deep neural networks?
• How are neural networks trained from scratch?

Fundamentals of neural networks with TensorFlow

• Regression
• Binary classification
• Multiclass classification
• Multi-label classification
• Activation functions
• Optimization

Training deep neural networks

• Embeddings
• Faster optimization
• Vanishing/exploding gradients and solutions
• Regularization

Analyzing sequence data

• Recurrent Neural Networks
• Attention

Analyzing images

• How does a computer store images?
• Convolutional neural networks

The course features several group assignments and a bring-your-own project for learners to further develop their understanding of the content. In addition to the group assignments, short individual assignments are offered every session. This course is ideal for learners who work full-time and want to begin working with deep learning or refresh their knowledge of it.

Basic knowledge of Python and statistics is required to start this course.