Machine Learning.AI

1. Definition of machine learning

Broadly speaking, machine learning is a method that can give a machine the ability to learn so that it can perform functions that cannot be performed by direct programming. But in a practical sense, machine learning is a method of using data, training a model, and then using the model to predict.

Let's look specifically at an example of housing prices.

Take the national topic house as an example. Now I have a house for sale, what price should I put on it? The area of the house is 100 square meters, and the price is 1 million, 1.2 million, or 1.4 million?

Obviously, I hope to obtain a certain law of house price and area. So how do I get this regularity? Use the average house price data in the newspaper? Or refer to others with similar areas? Either way, it doesn't seem very reliable.

Now I hope to obtain a reasonable law that can reflect the relationship between area and house price to the greatest extent. So I investigated some surrounding houses similar to mine and obtained a set of data. This set of data contains the area and price of large and small houses. If I can find out the law of area and price from this set of data, then I can get the price of the house.

The search for the regularity is very simple, a straight line is fitted, let it "pass through" all the points, and the distance from each point is as small as possible.

Through this straight line, I obtained a law that can best reflect the law of house price and area. This line is at the same time a function expressed by:

Price = Area * a + b

Both a and b in the above are parameters of the straight line. After obtaining these parameters, I can calculate the price of the house.

Suppose a = 0.75, b = 50, then house price = 100 * 0.75 + 50 = 1.25 million. This result is different from the 1 million, 1.2 million, and 1.4 million I listed earlier. Since this line takes most of the cases into account, it is the most reasonable prediction in a "statistical" sense.

Two pieces of information were revealed during the solution process:

1. The housing price model is determined according to the type of function fitted. If it is a straight line, then what is fitted is the equation of the line. If it is another type of line, such as a parabola, then what is fitted is the parabolic equation. There are many algorithms for machine learning, and some powerful algorithms can fit complex nonlinear models to reflect situations that cannot be expressed by straight lines.

2. The more data I have, the more situations my model can take into account, and the better the prediction effect on new situations may be. This is a manifestation of the "data is king" thinking in the machine learning community. Generally speaking (not absolutely), the more data, the better the prediction effect of the model generated by machine learning.

Through my process of fitting a straight line, we can do a complete review of the machine learning process. First, we need to store historical data in the computer. Next, we process these data through machine learning algorithms. This process is called "training" in machine learning, and the processed results can be used by us to predict new data. This result is generally called a "model". The process of making predictions on new data is called "prediction" in machine learning. "Training" and "prediction" are two processes of machine learning, "model" is the intermediate output of the process, "training" produces "model", and "model" guides "prediction".

2. The scope of machine learning

Machine learning has deep connections with pattern recognition, statistical learning, data mining, computer vision, speech recognition, natural language processing and other fields.

In terms of scope, machine learning is similar to pattern recognition, statistical learning, and data mining. At the same time, the combination of machine learning and processing technologies in other fields forms interdisciplinary subjects such as computer vision, speech recognition, and natural language processing. Therefore, when talking about data mining in general, it can be equivalent to talking about machine learning. At the same time, the machine learning applications we usually refer to should be general-purpose, not limited to structured data, but also image, audio and other applications.

The following figure shows some related disciplines and research fields involved in machine learning.

pattern recognition

Pattern recognition = machine learning. The main difference between the two is that the former is a concept developed from the industry, while the latter is mainly derived from computer science. In the famous book "Pattern Recognition And Machine Learning", Christopher M. Bishop said at the beginning that "pattern recognition comes from industry, and machine learning comes from computer science. However, the activities in them can be viewed as They are two sides of the same field, and they have both grown considerably in the past 10 years."

data mining

Data mining = machine learning + database. The concept of data mining in recent years is too familiar. Almost tantamount to hype. Anyone who talks about data mining will brag about how data mining is, such as digging out gold from data, and converting discarded data into value, and so on. However, although I may dig out gold, I may also dig out "stone". This statement means that data mining is just a way of thinking, telling us that we should try to mine knowledge from data, but not every data can mine gold, so don't myth it. A system will never become omnipotent because of a data mining module (this is what IBM likes to brag about most). On the contrary, a person with data mining thinking is the key, and he must also have a deep understanding of data. Only in this way can it be possible to derive patterns from the data to guide business improvement. Most of the algorithms in data mining are the optimization of machine learning algorithms in the database.

statistical learning

Statistical learning is approximately equal to machine learning. Statistical learning is a discipline that highly overlaps with machine learning. Because most methods in machine learning come from statistics, it can even be considered that the development of statistics promotes the prosperity of machine learning. For example, the famous support vector machine algorithm is derived from statistics. But there is a difference between the two to a certain extent. The difference is that statistical learners focus on the development and optimization of statistical models, and prefer mathematics, while machine learners pay more attention to problem solving and practice, so Machine learning researchers will focus on improving the efficiency and accuracy of learning algorithms executed on computers.

computer vision

Computer Vision = Image Processing + Machine Learning. Image processing techniques are used to process images into suitable inputs into machine learning models, and machine learning is responsible for identifying relevant patterns from images. There are many applications related to computer vision, such as Baidu image recognition, handwritten character recognition, license plate recognition and other applications. This field has a very hot application prospect and is also a hot research direction. With the development of deep learning in the new field of machine learning, the effect of computer image recognition has been greatly promoted, so the future development prospects of the computer vision industry are immeasurable.

Speech Recognition

Speech recognition = speech processing + machine learning. Speech recognition is the combination of audio processing technology and machine learning. Speech recognition technology is generally not used alone, and is generally combined with related technologies of natural language processing. Current related applications include Siri, Apple's voice assistant.

natural language processing

Natural language processing = text processing + machine learning. Natural language processing technology is mainly a field that allows machines to understand human language. In natural language processing technology, a large number of technologies related to compilation principles are used, such as lexical analysis, syntax analysis, etc. In addition, at the level of understanding, technologies such as semantic understanding and machine learning are used. As the only symbol created by humans, natural language processing has been the direction of continuous research in the machine learning community. According to Yu Kai, a Baidu machine learning expert, "Listening and reading, to put it bluntly, can be understood by cats and dogs, but only language is unique to human beings." How to use machine learning technology for in-depth understanding of natural language has always been the focus of industry and academia.

The next chapter talks about machine learning algorithms.