Building Regression Models Using TensorFlow

TensorFlow is the tool of choice for building deep learning applications. In this course, you'll learn how the neurons in neural networks learn non-linear functions, and how neural networks execute operations such as regression and classification.
Course info
Level
Intermediate
Updated
Jul 10, 2017
Duration
2h 40m
Table of contents
Description
Course info
Level
Intermediate
Updated
Jul 10, 2017
Duration
2h 40m
Description

TensorFlow is all about building neural networks that can "learn" functions, and linear regression can be learnt by the simplest possible neural network - of just 1 neuron! In contrast, the XOR function requires 3 neurons arranged in 2 layers, and smart image recognition can require thousands of neurons. In this course, Building Regression Models using TensorFlow, you'll learn how the neurons in neural networks learn non-linear functions. First, you'll begin by learning functions such as XOR, and how to train different gradient descent optimizers. Next, you'll dive into the implications of choosing activation functions, such as softmax and ReLU. Finally, you'll explore the use of built-in estimators in Tensorflow. By the end of this course, you'll have a better understanding of how neurons "learn", and how neural networks in TensorFlow are set up and trained to execute operations such as regression and classification.

About the author
About the author

An engineer and tinkerer, Vitthal has worked at Google, Credit Suisse, and Flipkart and studied at Stanford and INSEAD. He has worn many hats, each of which has involved writing code and building models. He is passionately devoted to his hobby of laughing at his own jokes.

More from the author
More courses by Vitthal Srinivasan
Section Introduction Transcripts
Section Introduction Transcripts

Course Overview
Hi, everyone. My name is Vitthal Srinivasan. Welcome to my course, Building Regression Models Using TensorFlow. I am co-founder at a startup named Loonycorn, and before this I've worked at Google and studied at Stanford. TensorFlow is all about building neural networks that can learn functions and linear regression can be learned by the simplest possible neural network, consisting of just one neuron. In contrast, the XR logical function requires at least three neurons arranged in two layers, and smart image recognition can require thousands of neurons. Some of the major topics that we will cover include how the neurons and neural networks can learn non-linear functions such as XR, the neural networks that are used for linear and logistic regression, training different gradient descent optimizers, the implications of activation functions such as softmax and ReLU, and the use of in-built estimators in TensorFlow to simplify regression. By the end of this course, you will have a good understanding of now neurons learn, and how neural networks in TensorFlow are set up and trained to execute operations such as regression and classification. I hope you'll join me on this journey to learn how to wire up neurons and neural networks in TensorFlow, with the course Building Regression Models Using TensorFlow at Pluralsight. (calm music)