Machine learning with TensorFlow: From linear algebra to neural networks

Machine learning capabilities in software or apps is no longer a “nice to have” feature, and the open source machine learning framework TensorFlow is becoming an industry standard. Rich Ott explores simple machine learning models such as classification and regression models, constructs and launches graphs in TensorFlow using TensorBoard to visualize workflow and build and test models in TensorFlow using real-world data. You’ll leave with both a theoretical and practical understanding of the algorithms behind machine learning and be ready to incorporate them into your next project.

You’ll be taught using TensorFlow’s Python interface.

Outline

Day 1

• Practical linear algebra
• Introduction to TensorFlow
• Iterative algorithms

Exercises

• Implementing a basic graph
• Reducing tensors of arbitrary shape
• Fibonacci numbers
• Minimizing functions

Day 2

• Machine learning
• Basic neural networks
• Deep neural networks

Exercises

• Multidimensional linear regression
• Tuning hyperparameters and visualizing the weight matrix
• Build an Iris classifier
• Adding neurons and layers to a neural network
• Implementing early stopping
• Exploring activation functions, dropout, and learning rates