Course Description
As companies continue to embrace digital tools many now find themselves with access to an unprecedented amount of data. While business leaders are aware of the importance of data as a resource, many still struggle with how to capitalize on its value. Deep learning techniques can help your data and analytics team find insights in a data lake that would take an individual decades to process. In Deep Learning Foundations by Metis, course attendees will receive a solid background in the topic as well as instruction on how to get started with the most common Python libraries for doing Deep Learning. This course is geared towards individuals who have some programming and machine learning experience.
Course Outcomes
Upon completion of the course, attendees should be able to:
Define deep learning and common terminology
Articulate the strengths and weaknesses of deep learning and identify the types of problems deep learning can solve
Explain how deep learning models are trained and practical techniques for training a deep learning model
Use a variety of preprocessing and initialization techniques
List, differentiate, and use different activation functions for deep learning models and the different optimizers to train them
And much more...
Training Content
Day 1
- Machine Learning Review
- Neural Networks Overview & Workflow
- Fully Connected Networks
- Activation Functions
- Loss Functions
- Training Neural Networks
- Practical Issues
- GPUs
Day 2
- Review Day 1
- More Activation Functions
- Optimizers
- Regularization
Day 3
- Review Day 2
- Introduction to CNNs
- Hyperparameters
- Data Augmentation
- Pre-trained Models
- Transfer Learning
- Network Architectures
Day 4
- Review Day 3
- RNNs
- RNN Extensions
- Applications
- Seq2Seq
- Application of Seq2Seq and Attention to Language Translation
Day 5
- Review of Day 4
- Word Vectors
- Unsupervised Representations of Images
- Application of Unsupervised Learning to Image Similarity
- GANs
Related Blog Posts
More Courses for Your Team
