Deep learning traditional machine

Some problems that human beings can solve quickly through intuition are difficult for computers to solve at present. These problems include natural speech understanding, image recognition, speech recognition and so on, all of which need to be solved by artificial intelligence.

Deep learning is a branch of machine learning, which can not only learn the relationship between features and tasks, but also automatically extract more complex features from simple features. Deep learning algorithm can learn more complex feature expressions from data, which makes the last step of weight learning simpler and more effective. Deep learning can gradually transform simple features into more complex features, thus making it easier to distinguish different types of images. The difference between deep learning and traditional machine learning is shown in the figure;

Artificial intelligence, machine learning and deep learning are very related fields. The following figure summarizes the relationship between them. Artificial intelligence is a very extensive problem of human beings, and machine learning is an important means to solve this kind of problem. Deep learning is a branch of machine learning. On many artificial intelligence problems, the method of deep learning has broken through the bottleneck of traditional machine learning methods and promoted the development of artificial intelligence.