Oreilly - Beginning Application Development with TensorFlow and Keras
by Nimish Narang, Luis Capelo | Released July 2018 | ISBN: 9781789343557
Learn to design, develop, train, and deploy TensorFlow and Keras models as real-world applicationsAbout This VideoProvides detailed explanation of neural networks and deep learningMake predictions with a trained model and get to grips with TensorBoardPerfectly balances theory, hands-on demos, and assessmentsIn DetailWith this course, you'll learn how to train, evaluate, and deploy Tensorflow and Keras models as real-world web applications. After a hands-on introduction to neural networks and deep learning, you'll use a sample model to explore details of deep learning and learn to select the right layers that can solve a given problem. By the end of the course, you'll build a Bitcoin application that predicts the future price, based on historic and freely available information. Show and hide more
- Chapter 1 : Introduction to Neural Networks and Deep Learning
- Course Overview 00:04:02
- Setting up your Environment 00:05:06
- Lesson Overview 00:01:36
- What are Neural Networks and Deep Learning? 00:06:37
- Limitations of Deep Learning 00:05:05
- Common Components and Operations of Neural Networks 00:06:04
- Configuring a Deep Learning Environment 00:01:23
- Installing Python 3 00:05:29
- Installing TensorFlow, Keras and TensorBoard 00:09:36
- Installing Jupyter, Notebooks, Pandas and NumPy 00:03:15
- Installation Completion 00:04:16
- Training a Neural Network with TensorFlow convolutional layer 00:05:07
- Training a Neural Network with TensorFlow fully connected layer 00:02:43
- Train a Neural Network with TensorFlow 00:06:07
- Testing network performance with unseen data 00:03:51
- Summary 00:01:23
- Chapter 2 : Model Architecture
- Lesson Overview 00:01:04
- Choosing the Right Model Architecture 00:08:39
- Data Normalization 00:05:25
- Using Keras as a TensorFlow Interface 00:02:38
- Designing a Model 00:05:11
- Training a Model 00:03:57
- Making Predictions 00:02:24
- The Keras Paradigm 00:02:10
- From Data Preparation to Modeling 00:05:34
- Reshaping the Time-Series Data 00:09:20
- Reshaping the Time-Series Data 00:03:52
- Training a Model 00:02:53
- Training a Model 00:01:17
- Making Predictions 00:01:53
- Overfitting 00:01:01
- Summary 00:00:44
- Chapter 3 : Model Evaluation and Optimization
- Lesson Overview 00:01:13
- Model Evaluation 00:10:15
- Using TensorBoard 00:08:58
- Implementing Model Evaluation Metrics 00:07:24
- Evaluating Bitcoin Model 00:14:19
- Model Predictions 00:11:55
- Interpreting Predictions 00:05:47
- Hyperparameter Optimization 00:09:49
- Epochs Implementation 00:10:50
- Regularization Strategies Implementation 00:06:48
- Summary 00:01:27
- Chapter 4 : Productization
- Lesson Overview 00:01:34
- Handling and Dealing with New Data 00:12:15
- Re-Training an Old Model 00:12:46
- Training a New Model 00:03:33
- Deploying a Model as a Web Application 00:04:20
- Building and executing a Docker run command 00:03:49
- Deployment and using Cryptonic 00:08:59
- Summary 00:01:38
Show and hide more