O'Reilly logo

Stay ahead with the world's most comprehensive technology and business learning platform.

With Safari, you learn the way you learn best. Get unlimited access to videos, live online training, learning paths, books, tutorials, and more.

Start Free Trial

No credit card required

TensorFlow and the Google Cloud ML Engine for Deep Learning

Video Description

CNNs, RNNs and other neural networks for unsupervised and supervised deep learning

About This Video

  • Deep learning basics: What a neuron is; how neural networks connect neurons to 'learn' complex functions; how TF makes it easy to build neural network models
  • Using Deep Learning for the famous ML problems: regression, classification, clustering and autoencoding
  • CNNs - Convolutional Neural Networks: Kernel functions, feature maps, CNNs v DNNs
  • RNNs - Recurrent Neural Networks: LSTMs, Back-propagation through time and dealing with vanishing/exploding gradients
  • Unsupervised learning techniques - Autoencoding, K-means clustering, PCA as autoencoding
  • Working with images
  • Working with documents and word embedding’s
  • Google Cloud ML Engine: Distributed training and prediction of TF models on the cloud
  • Working with TensorFlow estimators

In Detail

TensorFlow is quickly becoming the technology of choice for deep learning, because of how easy TF makes it to build powerful and sophisticated neural networks. The Google Cloud Platform is a great place to run TF models at scale, and perform distributed training and prediction. This is a comprehensive, from-the-basics course on TensorFlow and building neural networks. It assumes no prior knowledge of Tensorflow, all you need to know is basic Python programming.

Table of Contents

  1. Chapter 1 : Introduction
    1. You, This Course and Us 00:02:38
  2. Chapter 2 : Installation
    1. Install TensorFlow 00:06:24
    2. Install Jupyter Notebook 00:04:38
    3. Lab: Setting up A GCP Account 00:07:00
    4. Lab: Using the Cloud Shell 00:06:01
    5. Datalab ~ Jupyter 00:03:00
    6. Lab: Creating and Working On a Datalab Instance 00:04:02
  3. Chapter 3 : TensorFlow and Machine Learning
    1. Introducing Machine Learning 00:08:04
    2. Representation Learning 00:10:28
    3. Neural Networks Introduced 00:07:35
    4. Introducing TensorFlow 00:07:17
    5. Lab: Simple Math Operations 00:08:46
    6. Computation Graph 00:10:17
    7. Tensors 00:09:02
    8. Lab: Tensors 00:05:04
    9. Linear Regression Intro 00:09:57
    10. Placeholders and Variables 00:08:45
    11. Lab: Placeholders 00:06:37
    12. Lab: Variables 00:07:49
    13. Lab: Linear Regression with Made-up Data 00:04:52
  4. Chapter 4 : Working with Images
    1. Image Processing 00:08:06
    2. Images As Tensors 00:08:16
    3. Reading and Working with Images 00:08:06
    4. Lab: Image Transformations 00:06:38
  5. Chapter 5 : K-Nearest-Neighbors with TensorFlow
    1. Introducing MNIST 00:04:13
    2. K-Nearest Neighbours as Unsupervised Learning 00:07:43
    3. One-hot Notation and L1 Distance 00:07:31
    4. Steps in the K-Nearest-Neighbours Implementation 00:09:33
    5. Lab: K-Nearest-Neighbours 00:14:15
  6. Chapter 6 : Linear Regression with a Single Neuron
    1. Learning Algorithm 00:10:59
    2. Individual Neuron 00:09:52
    3. Learning Regression 00:07:51
    4. Learning XOR 00:10:27
    5. XOR Trained 00:11:12
  7. Chapter 7 : Linear Regression in TensorFlow
    1. Lab: Access Data from Yahoo Finance 00:02:50
    2. Non TensorFlow Regression 00:08:06
    3. Lab: Linear Regression - Setting Up a Baseline 00:11:19
    4. Gradient Descent 00:09:57
    5. Lab: Linear Regression 00:14:42
    6. Lab: Multiple Regression in TensorFlow 00:09:16
  8. Chapter 8 : Logistic Regression in TensorFlow
    1. Logistic Regression Introduced 00:10:16
    2. Linear Classification 00:05:26
    3. Lab: Logistic Regression - Setting Up a Baseline 00:07:33
    4. Logit 00:08:33
    5. Softmax 00:11:55
    6. Argmax 00:12:13
    7. Lab: Logistic Regression 00:16:57
  9. Chapter 9 : The Estimator API
    1. Estimators 00:04:11
    2. Lab: Linear Regression using Estimators 00:07:49
    3. Lab: Logistic Regression using Estimators 00:04:54
  10. Chapter 10 : Neural Networks and Deep Learning
    1. Traditional Machine Learning 00:06:24
    2. Deep Learning 00:09:23
    3. Operation of a Single Neuron 00:08:18
    4. The Activation Function 00:10:42
    5. Training a Neural Network: Back Propagation 00:06:41
    6. Lab: Automobile Price Prediction - Exploring the Dataset 00:11:13
    7. Lab: Automobile Price Prediction - Using TensorFlow for Prediction 00:14:35
    8. Hyperparameters 00:06:27
    9. Vanishing and Exploding Gradients 00:12:11
    10. The Bias-Variance Trade-off 00:08:27
    11. Preventing Overfitting 00:07:37
    12. Lab: Iris Flower Classification 00:12:09
  11. Chapter 11 : Classifiers and Classification
    1. Classification as an ML Problem 00:07:50
    2. Confusion Matrix: Accuracy, Precision and Recall 00:12:39
    3. Decision Thresholds and the Precision-Recall Trade-off 00:10:44
    4. F1 Scores and the ROC Curve 00:07:45
  12. Chapter 12 : Convolutional Neural Networks (CNNs)
    1. Mimicking the Visual Cortex 00:05:08
    2. Convolution 00:06:43
    3. Choice of Kernel Functions 00:04:48
    4. Zero Padding and Stride Size 00:05:48
    5. CNNs vs DNNs 00:07:15
    6. Feature Maps 00:09:29
    7. Pooling 00:06:15
    8. Lab: Classification of Street View House Numbers - Exploring the Dataset 00:10:38
    9. Basic Architecture of a CNN 00:07:07
    10. Lab: Classification of Street View House Numbers - Building the Model 00:12:53
    11. Lab: Classification of Street View House Numbers - Running the Model 00:07:36
    12. Lab: Building a CNN Using the Estimator API 00:12:20
  13. Chapter 13 : Recurrent Neural Networks (RNNs)
    1. Learning from the Past 00:08:32
    2. Unrolling an RNN Cell through Time 00:06:55
    3. Training an RNN - Back Propagation through Time 00:08:23
    4. Lab: RNNs for Image Classification 00:14:21
    5. Vanishing and Exploding Gradients in an RNN 00:07:05
    6. Long Memory Neurons vs Truncated BPTT 00:06:03
    7. The Long/Short Term Memory Cell 00:06:29
    8. A Sequence of Words 00:06:36
    9. Text in Numeric Form 00:15:08
    10. Lab: Sentiment Analysis on Rotten Tomatoes Reviews - Exploring the Dataset 00:10:35
    11. Lab: Sentiment Analysis on Rotten Tomatoes Reviews - Building, Running the Model 00:11:20
  14. Chapter 14 : Unsupervised Learning
    1. Supervised and Unsupervised Learning 00:11:30
    2. Expressing Attributes as Numbers 00:05:33
    3. K-Means Clustering 00:15:15
    4. Lab: K-Means Clustering with 2-Dimensional Points in Space 00:08:52
    5. Lab: K-Means Clustering with Images 00:10:20
    6. Patterns in Data 00:03:20
    7. Principal Components Analysis 00:13:20
    8. Autoencoders 00:05:03
    9. Autoencoder Neural Network Architecture 00:09:04
    10. Lab: PCA on Stock Data - Matplotlib vs Autoencoders 00:14:15
    11. Stacked Autoencoders 00:04:28
    12. Lab: Stacked Autoencoder with Dropout 00:07:52
    13. Lab: Stacked Autoencoder with Regularization and He Initialization 00:06:15
    14. Denoising Autoencoders 00:01:26
    15. Lab: Denoising Autoencoder with Gaussian Noise 00:01:58
  15. Chapter 15 : TensorFlow on the Google Cloud
    1. Lab: Taxicab Prediction - Setting up the dataset 00:14:39
    2. Lab: Taxicab Prediction - Training and Running the model 00:11:22
  16. Chapter 16 : TensorFlow Using Cloud ML Engine
    1. A Taxicab Fare Prediction Problem 00:03:25
    2. Datalab 00:07:03
    3. Querying BigQuery 00:05:24
    4. Explore Data 00:06:03
    5. Clean Data 00:04:48
    6. Benchmark 00:05:45
    7. Using TensorFlow 00:08:22
    8. The Estimator API 00:08:48
    9. The Experiment Function 00:05:48
    10. Introduction to Cloud MLE 00:07:53
    11. Using Cloud MLE 00:08:06
    12. The Training Service 00:06:24
    13. The Prediction Service 00:07:54
  17. Chapter 17 : Feature Engineering and Hyperparameter Tuning
    1. Feature Engineering to the rescue 00:01:04
    2. New Approach 00:06:43
    3. Dataflow Create Pipeline 00:07:11
    4. Dataflow Run Pipeline 00:05:04
    5. Feature Engineering 00:08:35
    6. Deep and Wide Models 00:09:16
    7. Hyperparameter Tuning 00:07:34
    8. Hyperparameter Tuning on the GCP 00:06:36