This is the code repository for Deep Learning Architectures and Applications [Video], published by Packt. It contains all the supporting project files necessary to work through the video course from start to finish.
This video course presents deep learning architectures coded in Python using Keras, a modular neural network library that runs on top of either Google's TensorFlow or Lisa Lab's Theano backends. This video course introduces Generative Adversarial Networks (GANs) that are used to reproduce synthetic data that looks like data generated by humans, and then teach how to forge the MNIST and CIFAR-10 dataset with the help of Keras Adversarial GANs.
Practical applications include code for predicting the surrounding words given the current word, sentiment analysis, and synthetic generation of texts. We will learn about a specific form of word embedding word2vec. This embedding has proven more effective and has been widely adopted in the deep learning and NLP communities. We will also learn different ways in which you can generate your own embeddings in your Keras code.
By the end of this video course, you will be able to transform words in text into vector embeddings that retain the distributional semantics of the word.
- Grasp the Concepts and intuitive understating of Deep Learning
- Build your own Multilayer Neural Networks
- Explore the Concepts of Convolutional Neural Networks and Recurrent Neural Networks
- Build Convolutional Neural Networks and Recurrent Neural Networks
- Use transfer learning to greatly increase CNN performance
- Use Concepts, intuitive understating and applications of Autoencoders and Generative Adversarial Networks
- Build Auto encoders and Generative Adversarial Networks
This course is a concise yet thorough introduction to GAN, RNN, word embeddings and deep learning technologies designed especially for software engineers and data scientists. Knowledge of Python is required for this course.
This course has the following software requirements:
To be able to smoothly follow through the sections, you will need the following pieces of software:
TensorFlow 1.0.0 or higher
Keras 2.0.2 or higher
Matplotlib 1.5.3 or higher
Scikit-learn 0.18.1 or higher
NumPy 1.12.1 or higher
The hardware specifications are as follows:
Either 32-bit or 64-bit architecture
2+ GHz CPU
4 GB RAM
At least 10 GB of hard disk space available