 Название: Artificial Intelligence: Evolution and Revolution
Автор: Steven Astorino, Mark Simmonds
Издательство: MC Press
Год: 2019
Страниц: 121
Язык: английский
Формат: epub, rtf, pdf (conv)
Размер: 10.1 MB
From humble evolutions in research papers and labs, artificial intelligence (AI) - which encompasses Machine Learning (ML) and Deep Learning (DL) - has matured in its many forms, infused in applications that can learn on their own and become progressively smarter with each interaction and transaction. Coupled with immense stores of data, maturity in CPU and GPU hardware, the invention of new, open source deep learning algorithms, and cloud technologies, operational AI has become available to the masses, setting the wheels in motion for a worldwide AI revolution that has never been seen before. This book attempts to help the reader on their AI journey by covering the concepts of AI, Machine Learning, and Deep Learning in its many forms; key ML and DL algorithms data scientists should learn; ethical challenges for the use of AI; how AI is being used across industries; possible future outlook for AI, and an AI Ladder to help accelerate the AI journey.
Go from Machine Learning to Deep Learning (DL) within Watson Studio, a cloud-based, experiment-centric model-training environment offering a powerful set of tools and pay-as-you-use resources. It includes Neural Network Modeler, an intuitive drag-and-drop interface for designing neural architectures using the most popular deep-learning frameworks: TensorFlow, Caffe, PyTorch, and Keras. And users can quickly capture network designs and then export them for experimental optimization.
Experiment Builder is a suite of tools that guide and manage the experimental process, simplifying the hundreds to thousands of training runs it takes to find the right combination of network layer configurations and hyperparameters. Each training run is automatically started, monitored, and stopped upon completion. Training progress and cross-model performance can be viewed in real time and revisited later. Training history and assets are tracked and then automatically transferred to object storage.
Скачать Artificial Intelligence: Evolution and Revolution
|
Разделы сайта
Авторизация
Реклама
Artificial Intelligence: Evolution and Revolution 
