Embrace artificial intelligence, start with machine learning

Intelligence is a very common word in modern life, such as smart phones, smart home products, intelligent robots, etc., but the meaning of intelligence is different in different occasions. What we call "Artificial Intelligence" (AI) is to let the machine think like a human being and have human intelligence.

Since its birth, the field of artificial intelligence has experienced prosperity and lows again and again. Its development can be roughly divided into "inference period", "knowledge period" and "learning period". The reasoning period mainly focuses on logical reasoning but the perceptron is too simple; although various knowledge systems are established in the knowledge period, the self-learning ability and the neural network resource capacity are insufficient. During the learning period, the machine was able to learn the knowledge itself, and it was not until 1980 that machine learning became a popular subject because of its outstanding performance in many fields. In modern times, with the development of the Internet, cloud computing, big data, and the enhancement of GPU, chip and software technology, deep learning has begun to expand, expanding the scope of artificial intelligence, and also promoting the transformation of society from digital to intelligent.

Artificial intelligence mainly consists of several parts: first, perception, including vision, speech, language; then decision making, such as making predictions and judgments; and finally feedback, if you want to do a complete system, like robot or automatic Driving, you need a feedback.

One of the most important abilities of artificial intelligence is its learning ability - machine learning, which is the core of artificial intelligence and the key to making computers intelligent. I can't learn by myself, and artificial intelligence is just a matter of words.

To understand artificial intelligence, we need to clarify the relationship between several concepts: artificial intelligence is a big concept, let the machine think like humans and even surpass humans; and machine learning is a way to realize artificial intelligence, using algorithms To analyze data, learn from it, and then make decisions and predictions on events in the real world; deep learning is an implementation of machine learning, training the network by simulating human neural networks; and statistics is machine learning and neural A basic knowledge of the network.

The biggest feature of machine learning is the use of data rather than instructions to carry out various tasks. The learning process mainly includes: data feature extraction, data preprocessing, training model, test model, model evaluation and improvement. Next we focus on the common algorithms in the machine learning process.