Keras 2.x Projects

9 projects demonstrating faster experimentation of neural network and deep learning applications using Keras

Giuseppe Ciaburro

• Editore: Packt Publishing
• Anno: 2018
• ISBN:
  9781789534160
• Edizione: 1
• Argomento: Ciencia de la computación
• Categoria: Inteligencia artificial (IA) y semántica

Demonstrate fundamentals of Deep Learning and neural network methodologies using Keras 2.x

Key Features

• Experimental projects showcasing the implementation of high-performance deep learning models with Keras.
• Use-cases across reinforcement learning, natural language processing, GANs and computer vision.
• Build strong fundamentals of Keras in the area of deep learning and artificial intelligence.

Book Description

Keras 2.x Projects explains how to leverage the power of Keras to build and train state-of-the-art deep learning models through a series of practical projects that look at a range of real-world application areas.

To begin with, you will quickly set up a deep learning environment by installing the Keras library. Through each of the projects, you will explore and learn the advanced concepts of deep learning and will learn how to compute and run your deep learning models using the advanced offerings of Keras. You will train fully-connected multilayer networks, convolutional neural networks, recurrent neural networks, autoencoders and generative adversarial networks using real-world training datasets. The projects you will undertake are all based on real-world scenarios of all complexity levels, covering topics such as language recognition, stock volatility, energy consumption prediction, faster object classification for self-driving vehicles, and more.

By the end of this book, you will be well versed with deep learning and its implementation with Keras. You will have all the knowledge you need to train your own deep learning models to solve different kinds of problems.

What you will learn

• Apply regression methods to your data and understand how the regression algorithm works
• Understand the basic concepts of classification methods and how to implement them in the Keras environment
• Import and organize data for neural network classification analysis
• Learn about the role of rectified linear units in the Keras network architecture
• Implement a recurrent neural network to classify the sentiment of sentences from movie reviews
• Set the embedding layer and the tensor sizes of a network

Who this book is for

If you are a data scientist, machine learning engineer, deep learning practitioner or an AI engineer who wants to build speedy intelligent applications with minimal lines of codes, then this book is the best fit for you. Sound knowledge of machine learning and basic familiarity with Keras library would be useful.

Reuters Newswire Topics Classifier in Keras

Natural language processing (NLP) is the process of automatic processing of information written or spoken in a natural language using an electronic calculator. This is made particularly difficult and complex due to the intrinsic characteristics of the ambiguity of human language. When it's necessary to make the machine learn methods of interaction with the environment typical of man, the question isn't so much that of storing data, but that of letting the machine learn how this data can be translated simultaneously to create a concept. Natural language interacts with the environment generating predictive knowledge.

In this chapter, Keras layers are used to build a model to classify Reuter's newswire topics. Data is available from a dataset that contains 11,228 newswires from Reuters with 46 labeled topics. This dataset is from Keras, and each wire is encoded as a sequence of word indexes.

The following topics will be covered in this chapter:

• Natural language processing
• The Natural Language Toolkit
• Corpora
• Word and sentence tokenize
• Implementing a DNN to label sentences

In this chapter, we'll learn how to analyze, understand, and generate texts using Natural Language Toolkit. Then, we will analyze the most frequent applications using natural language processing. We'll learn about Keras sequential layers and understand training and validation procedures. Finally, we will explore a practical case: how to label sentences contained in the Reuters newswire topics dataset using a Keras Deep Neural Network.

Natural language processing

NLP aims to implement IT tools to analyze, understand, and generate texts that people can understand naturally, as if they were communicating with another human interlocutor and not a computer. By natural language, we mean the language that we use in everyday life, such as English, Chinese, or Arabic, and that's synonymous with human language, mainly to distinguish it from formal language, including computer language. Natural language is the most natural and common form of communication, not only in its spoken version, but also in its written one. Compared to formal language, natural language is much more complex and often contains implications and ambiguities, which makes it very difficult to elaborate. Two goals can be pursued: text analysis and text generation. These characteristics define the following disciplines:

• Natural language analysis (NLA): This aims to study the understanding of natural language by machines. Deals with the reduction of a text in natural language in a formal and univocal representation.
• Natural language generation (NLG): This aims to study how machines can generate sentences in natural language. It deals with the construction of applications capable of producing correct sentences in a specific language.

These activities have a certain complexity and require different types of skills ranging from computer science to linguistics.

NLP phases

NLP involves several phases in the text analysis that make it a layered structure. The fundamental levels on which the analysis of a sentence is based are listed in the following:

• Morphology analysis
• Syntax analysis
• Semantic analysis
• Pragmatic analysis

Let's understand in depth about each layer.

Morphology analysis

The purpose of this phase is to subdivide the language input strings into sets of tokens corresponding to discrete words, sub-words, and punctuation forms. The text is then fragmented through a tokenization process, forming a series of tokens—each token is associated with a word of the text. In this phase, two terms assume relevance: stem and lemma. The stem is the root of a word, obtainable by removing from its flexed form (for example the conjugation of a verb or the plural of a noun) the ending. The lemma is instead the canonical form of the word, which is the one that's conventionally chosen to represent all of the inflected forms.

This phase has the task of identifying the stem and lemma of each word through two different operations, stemming and lemmatization respectively. This information will then be used in the subsequent analysis phases; in fact, it's much less expensive, in terms of memory, to maintain rules based on the parts that make up a word and how their combinations go to form certain inflected forms, rather than managing each word as an atomic element within a huge inventory.

Syntax analysis

To understand the meaning of a sentence, it's not enough to know the meaning of the words that compose it, but it's also essential to know how they relate to each other. This phase deals with the syntact...

Indice dei contenuti

• Title Page
• Copyright and Credits
• About Packt
• Contributors
• Preface
• Getting Started with Keras
• Modeling Real Estate Using Regression Analysis
• Heart Disease Classification with Neural Networks
• Concrete Quality Prediction Using Deep Neural Networks
• Fashion Article Recognition Using Convolutional Neural Networks
• Movie Reviews Sentiment Analysis Using Recurrent Neural Networks
• Stock Volatility Forecasting Using Long Short-Term Memory
• Reconstruction of Handwritten Digit Images Using Autoencoders
• Robot Control System Using Deep Reinforcement Learning
• Reuters Newswire Topics Classifier in Keras
• What is Next?
• Other Books You May Enjoy