Titlebook: Deep Learning and Practice with MindSpore; Lei Chen Book 2021 Tsinghua University Press 2021 Deep Learning.MindSpore.Deep Neural Networks

CHOIR

细胞膜

ODIUM

Incorporating Artificial IntelligenceThis chapter starts by outlining the historical development trends of AI and then explains what deep learning is and how it performs in practical applications. The chapter concludes by briefly describing the features of MindSpore, Huawei‘s self-developed deep learning framework.

NICE

Designing Human-Centric AI ExperiencesThis chapter describes several commonly used algorithms and basic concepts related to deep learning.

enormous

Working Effectively with AI Tech TeamsThis chapter introduces several important concepts related to DNN and presents some examples of using MindSpore to implement simple neural networks.

patriot

Popular Fallacies about HypertextThis chapter starts by describing the main challenges that face deep learning systems. It then explores the fundamentals involved in the training of DNNs, and concludes with some examples of using MindSpore to implement DNNs.

audiologist

Armando J. Oliveira,Duarte Costa PereiraIn this chapter, we describe the CNN. This network is a special neural network that uses convolution instead of general matrix multiplication at one or more layers. In essence, it is a feedforward neural network that uses convolutional mathematical operations.

extinct

Terry Mayes,Mike Kibby,Tony AndersonCalculations performed in the CNN are independent, meaning that there is no relationship between the previous and current inputs.

地名表

Studies in Systems, Decision and ControlVariational autoencoders (VAEs), proposed by Kingma and Welling in 2013, allow us to design complex generative models of data, which can then be trained to generate fictional images such as celebrity faces and high-resolution digital artworks.

暗语

Guidong Zhang,Bo Zhang,Zhong LiThis chapter starts by covering the basic concepts involved in reinforcement learning and then describes how to solve reinforcement learning tasks by using basic and deep learning-based solutions. It also provides a brief overview of the typical algorithms central to the deep learning-based solutions, namely DQN, DDPG, and A3C.