Authors: Kathrin Melcher and Rosaria Silipo (KNIME).
KNIME Analytics Platform is open-source software used to create and design data science workflows. This book is a comprehensive guide to KNIME and will enable you to integrate with various deep learning libraries to build neural network models without writing any code.
The book begins with an easy introduction to KNIME Analytics Platform, covering traditional feed-forward neural networks, and then shows you how to use a backpropagation algorithm with the help of practical examples. You’ll also learn how to build simple and more complex neural networks within KNIME Analytics Platform, without using a single line of code. You will start with a simple feed-forward network to solve a simple classification problem on a small dataset. Having covered the basic concepts, you’ll move on to prepare data accordingly; apply best practices to avoid overfitting; and build, train, test, and deploy more complex networks such as autoencoders, recurrent neural networks (RNNs), long short-term memory (LSTM), and convolutional neural networks (CNNs). In the concluding chapters, you’ll explore practical and creative solutions for solving real-world data problems.
By the end of the book, you’ll have learned how to build a number of different neural architectures and will be able to train, test, and deploy the network.
Publisher: Packt Publishing (December 9, 2020). Full details on Amazon.
A free copy of chapter one is available here for download.
Who the book is for
This book is for data analysts, data scientists, and deep learning developers who are not well versed with Python but want to learn how to use KNIME GUI to build, train, test, and deploy neural networks with different architectures. The practical implementations shown in the book do not require coding or any knowledge of dedicated scripts, allowing you to quickly translate theoretical concepts into practical applications. No previous knowledge of KNIME is required.
What you'll learn
- Use different common nodes to transform data into the right form to train a neural network
- Understand neural network techniques such as loss functions, backpropagation, and hyperparameters
- Prepare and encode data appropriately to feed the network and build a feed-forward neural architecture
- Build and train an autoencoder network to reproduce an input vector at the output layer
- Implement neural networks such as CNNs, RNNs, and LSTM with the help of practical examples
- Deploy a trained deep learning network on real-world data