1.2 What is Machine Learning
▲ What is Machine Learning?
Prior to figure out this problem, first find out what is learning.
Learning can be a human or animal by observing and thinking to get some skills course.
The machine learning Similarly, through computer data and calculated to obtain certain skills in the process.
Note that this comparison, machine learning is learning by observation through a data (a computer observation).
▲ then followed is to solve a new term appearing in the above "trick" (skill).
What is it skill? The trick is the ability of some of the more outstanding performance.
Machine learning techniques such as forecasting (prediction), identification (recognition).
To an example: to obtain an increase in revenue from data stock this technique, which is an example of a machine learning.
Now that people can get a skill through observation, why do you need a machine to learn it?
This is why machine learning, simply put, is for two reasons:
Some data or information, people can not get to, things that some people may be unrecognized, particularly large amount of information or data;
Another reason is that a process can not meet the needs of people, such as: definition of a lot of object recognition rule is satisfied, or other requirements; judgment in a short time by a large amount of information and the like.
The above said is why the use of machine learning, then under what circumstances the use of machine learning it? Is not all cases using machine learning it?
Here are three key elements of ML (machine learning abbreviation) of:
1, there is a pattern or performance allows us to improve it increase;
2, the rules are not so easily defined;
3, the need for data.
1.3 Applications of Machine Learning
Application of machine learning.
1.5 Machine Learning and Other Fields
Relationships with each other in the field of machine learning.
1.5.1 ML VS DM （Data Mining）
Machine learning and data mining are called Knowledge Discovery (KDD Knowledge Discovery in Dataset).
On a concept has been presented in machine learning, and therefore only the concepts under data mining is to find useful information from large amounts of data.
Starting from the definition, we can be the relationship between the two is divided into three kinds.
-
The two are the same: to be able to find useful information is obtained assuming that we require approximate objective function.
-
The two are mutual: the ability to find useful information can help us identify similar assumptions, and vice versa feasible.
-
Traditional data mining and computational problems more attention from a lot of data.
The total time, the two are inseparable.
1.5.2 M L VS AI （artificial intelligence）
Machine learning and artificial intelligence.
Artificial intelligence computer concept is probably able to show some wisdom behavior.
Can be obtained from the definition of machine learning is a way to achieve artificial intelligence.
1.5.3 ML VS statistic
Machine learning and statistics.
Statistics also need data to do an unknown inference.
Therefore, statistical machine learning method is an implementation.
The traditional statistical learning more concerned with the math formula, rather than the calculation itself.