Getting Started with TensorFlow 2.0 for Deep Learning

Name: Getting Started with TensorFlow 2.0 for Deep Learning
Rating: 3.5 (2 reviews)

Learn to develop deep learning models and kickstart your career in deep learning with TensorFlow 2.0

Created byPackt Publishing

Last updated 12/2019

English

English [Auto],

What you'll learn

Develop real-world deep learning applications
Classify IMDb Movie Reviews using Binary Classification Model
Build a model to classify news with multi-label
Train your deep learning model to predict house prices
Understand the whole package: prepare a dataset, build the deep learning model, and validate results
Understand the working of Recurrent Neural Networks and LSTM with hands-on examples
Implement autoencoders and denoise autoencoders in a project to regenerate images

Course content

7 sections • 35 lectures • 1h 46m total length

The Course Overview2:09
This video will give you an overview about the course.
Introducing Deep Learning and Setting Up the Environment5:06
Learn the basic concepts including AI, ML, and DL. Also, learn neural networks and perceptron.
• Distinguish between AI, ML, and DL
• Learn about NN
• Perceptron versus multilayer perceptron
Introducing the Universal Workflow of Deep Learning3:59
Define the universal workflow for any Deep Learning project.
• Introduce seven steps for any Deep Learning project
• Define the problem, get a dataset, and prepare the datasets
• Choose success metrics, evaluation criteria, and hyperparameter tuning
Mathematics Refresher4:02
Get familiar with the maths behind Deep Learning.
• Learn about tensors in Python
• Practical implementation
• Summarize the concepts
Training, Validation, and Test Sets1:24
Explore what train, validation, and test set are in a Deep Learning project.
• Define the train, validation, and test sets in the simplest of terms
• Download and manipulate Keras dataset
• Split the dataset into three sets programmatically
Data Preprocessing and Feature Engineering for Deep Learning Models1:48
Learn what data preprocessing and feature engineering are.
• Learn about vectorization, normalization, and handling missing values
• Learn about feature engineering
Demonstrating Overfitting and Underfitting of Data1:29
Distinguish between overfitting and underfitting.
• Explore the definitions of the core concepts of Deep Learning
• Define the goal of Deep Learning engineers
• Demonstrate with graphs
• 2.1 TensorFlow 2.0 Benefits and New Features
Test Your Knowledge

TensorFlow 2.0 Benefits and New Features3:04
Learn what is new in TensorFlow 2.0.
Explore PyTorch vs TensorFlow
Explore PyTorch vs TensorFlow 2.0
Learn about Keras in TensorFlow 2.0
Neural Networks in TensorFlow 2.03:37
Create a neural network in TensorFlow.
Import libraries
Get and preprocess datasets
Train and evaluate the trained model
Getting and Preprocessing MNIST Dataset2:44
Get, preprocess, and train a model for MNIST dataset.
Import libraries
Get and preprocess MNIST datasets
Train and evaluate the trained model
Building Neural Network for Classifying CIFAR10 Dataset in Jupyter Notebooks2:42
Get, preprocess, and train a model for CIFAR10 dataset.
Import libraries
Get and preprocess CIFAR10 datasets
Train and evaluate the trained model
TensorBoard and Callbacks2:23
Display loss and accuracy metrics in TensorBoard for MNIST training.
Learn about TensorBoard
Pass TensorBoard object as callback
Visualize the TF Graph and metrics
Test Your Knowledge

Introduction to Convolutional Neural Networks2:34
Introduce CNN.
• Define the convolutional layer
• Understand the pooling layer
• Know the dense layer
Loading the Dataset and Making Train and Test Sets3:48
Write a reusable script to load the entire dataset from videos.
Load and shuffle the dataset
Split the dataset into train and test sets
Dataset Preprocessing2:37
Learn about preprocessing.
• Get hands-on with preprocessing, before building the model
• Denoise the images in the dataset for better model training
• Resize the images to match the input layer of the model
Building CNN Model from Scratch with the Home Objects Dataset3:34
Build the CNN from ground up.
Define the convolutional layers
Summarize and compile the model
Train and evaluate the model
Data Augmentation to Overcome Overfitting and Generate Similar Images2:17
Learn a cool technique to avoid overfitting in almost any vision project.
• Learn about overfitting and data augmentation
• Use TensorFlow.keras functions to augment dataset
Test Your Knowledge

Introducing LSTM and Recurrent Neural Networks1:47
Introduction to RNN and LSTM.
• Know what a Recurrent Neural Network is
• Understand LSTM
Downloading and Working Textual Data: N-Grams and Bag-Of-Words2:29
Learn about N-gram, bag-of-words, and word embeddings.
• Understand N-grams
• Understand bag-of-words
• Understand word embeddings
Using GloVe Word Embeddings1:26
Learn about GloVe.
• Explore and understand GloVe
Building an LSTM Model for Text Data and Getting the Results4:57
Train RNN and LSTM from scratch.
Download IMDB datasets
Train with SimpleRNN
Train with LSTM
Test Your Knowledge

Introducing Recurrent Neural Networks2:26
Learn about LSTM, GRU, and Bidirectional Neural Network.
• Define RNNs
• Learn RNN architectures
• Know RNN applications
Getting and Preprocessing the Temperature Data3:59
Get and preprocess the temperature data.
• Download and preprocess the dataset with NumPy
• Visualize the entire dataset and temperature for ten consecutive days
• Write generators for train, validation, and test sets
Building an RNN Model for Temperature Forecasting2:57
Build a simple DL and GRU based model, for temperature prediction.
Develop a simple Deep Learning model
Visualize the results
Build the RNN with GRU layers and visualize the results
Test Your Knowledge

What Are Autoencoders? Why Are They Used?2:38
Learn about autoencoders.
• Explore autoencoders versus other dimensionality reduction techniques
• Understand the different types of autoencoders
• Know the autoencoder applications
Creating Simple Autoencoders from Scratch with the Fashion-MNIST Dataset3:02
Learn and develop simple autoencoders from scratch.
Preprocess the dataset
Build an autoencoder model
Train the model and visualize the results
Going in Depth with Deep Autoencoders2:48
Learn and develop deep autoencoders from scratch.
• Preprocess the dataset
• Build an autoencoder model
• Train the model and visualize the results
Reconstruct Images with Denoising Autoencoders1:42
Learn and develop simple autoencoders to denoise images.
• Preprocess the dataset
• Build an autoencoder model
• Train the model and visualize the results
Test Your Knowledge

TensorFlow 2.0 High Level API and Keras-API4:05
Get introduced to TensorFlow-Keras high level functional API.
• Get introduced to functional API
• Sample class creation with functional API
Getting and Preprocessing IMDB Dataset for Binary Classification5:10
Get and preprocess IMDB dataset for IMDB movie reviews.
• Get introduced to the problem and load the dataset
• Learn how to vectorize the entire dataset
• Preprocess and decode the encoded data points
Dataset Preparation, Building, and Validating Results on the IMDB Dataset4:59
Prepare the dataset, build the network, and validate the results.
Train the model
Validate the trained model
Display the history metrics
Getting and Preprocessing Reuters Dataset for Multi-Label Classification3:06
Get and preprocess Reuters dataset for news text.
• Get introduced to the problem and load the dataset
• Learn how to vectorize the entire dataset using one-hot-encoding
• Preprocess and decode the encoded data points
Dataset Preparation, Building, and Validating Results on the Reuters Dataset3:36
Prepare the dataset, build the network, and validate results.
Train the model
Validate the trained model
Display the history metrics
Getting and Preprocessing Dataset for Regression Analysis3:43
Get and preprocess Boston Housing Price dataset for news text.
• Get introduced to the problem and load the dataset
• Learn how to vectorize the entire dataset
Dataset Preparation on the Boston Housing Price Dataset2:14
Prepare the dataset, build the network, and validate results.
Train the model
Validate the trained model
Display the history metrics
Test Your Knowledge

Requirements

Basics of machine learning and now want to build deep learning systems with TensorFlow 2.0.

Description

Deep learning is a trending technology if you want to break into cutting-edge AI and solve real-world, data-driven problems. Google’s TensorFlow is a popular library for implementing deep learning algorithms because of its rapid developments and commercial deployments.

This course provides you with the core of deep learning using TensorFlow 2.0. You’ll learn to train your deep learning networks from scratch, pre-process and split your datasets, train deep learning models for real-world applications, and validate the accuracy of your models.

By the end of the course, you’ll have a profound knowledge of how you can leverage TensorFlow 2.0 to build real-world applications without much effort.

About the Author

Muhammad Hamza Javed is a self-taught machine learning engineer, an entrepreneur, and an author with over five years of industrial experience. Along with his team, he has been working on several computer vision, machine learning, and deep learning international projects. He learned skills on his own without a direct mentor, so he knows how troublesome it is for everyone to find to-the-point content that improves one’s skillset. He’s designed this course considering the challenges he faced when he learned and in projects, so you don’t have to spend too much time finding what’s best for you.

Who this course is for:

This course is for developers who have a basic knowledge of Python.
If you’re aware of the basics of machine learning and now want to build deep learning systems with TensorFlow 2.0 that are smarter, faster, more complex, and more practical, then this course is for you!

Getting Started with TensorFlow 2.0 for Deep Learning

What you'll learn

Explore related topics

Course content

Deep Learning Basics7 lectures • 20min

TensorFlow 2.0 for Deep Learning5 lectures • 15min

Working with CNNs for Computer Vision and Deep Learning5 lectures • 15min

Working with LSTM for Text Data and Deep Learning4 lectures • 11min

Working with RNNs for Time Series Sequences and Deep Learning3 lectures • 9min

Autoencoders – AE and Denoising AE4 lectures • 10min

Deep Learning Mini-Projects7 lectures • 27min

Requirements

Description

Who this course is for: