Author: Zen and the Art of Computer Programming
1 Introduction
2017 is a very important year. This year, the computer field developed vigorously, with the rapid development of new technologies such as AI, GAN, and VR. Coupled with the popularity of the mobile Internet, it has attracted great attention in various industries such as e-commerce, e-government, finance and insurance, government information, and security. Among them, deep learning (deep learning) is more and more widely used in various industries. This article will start from the perspective of deep learning, conduct a comprehensive analysis of it, and discuss its significance and limitations in practical applications.
In the second half of 2017, with the rise of deep learning in many fields such as medical treatment, image recognition, and natural language processing, deep learning is playing an increasingly important role in solving practical problems. Therefore, for deep learning, there are more and more problems to be solved, and its model structure, optimization strategy, training data volume and other factors are constantly changing. In order to better understand deep learning, you need to have a basic understanding of related concepts and theories.
2. Explanation of basic concepts and terms
deep learning
Deep learning refers to a kind of machine learning method. It uses a multi-level artificial neural network (Artificial Neural Network, ANN) to learn the feature representation of data by repeatedly superimposing and combining complex models composed of simple neurons. The characteristics of deep learning are:
- Models can automatically extract complex patterns in data.
- It can make full use of the correlation between the samples of the training data set and effectively predict the unknown data.
- Can handle large-scale data efficiently.
multi-task learning
Multitask learning refers to training multiple