Hi, boys and girls. this time we’re gonna talk about and the unsupervised. learning.
Ok. Now here is the question I want to ask you that what is the definition.
So it’s very simple for the definition. The last sentence we talk about the supervised learning that we have the answers, we have the standard to measure. What is the right what is wrong that we use the algorithm or the motor is the supervised learning? But for the unsupervised learning, we didn’t know what it is and we didn’t know which we do is Right and what we do wrong. If that means we didn’t have the answers about the simple or the raw data.
And I summarize the three kinds of items for your reference. The first one is according to the date feature, we want to make the connections in a connection about the data of the mode.
And for a lot of users who want to picture them out by different classification, which one wants to buy Which. things on the internet? And for all of you common sense JinRi TouTiao also used big data and machine learning technology to spread the news out for everyone who is interested in the tags, this tag was chosen by the machine learning algorithm.
The 3rd one is the connection rules, which can be expressed as the one and the baby user. This case is spread out for the machine learning the connection, the big data which is used as an example to express what happened in around us.
I want to emphasize more is. The other example or other useful technology in unsupervised learning is reinforcement learning.
This is the technology who is based on the reaction about the environment or the human body of the items wrong us based on their reaction, we used a different algorithm to choose the best next action based on the environment change. Of course, I think being this algorithm, they will use their learning by their self and have a decision based on the experiences they learned from the old world history. Generally, you can think that chicken games, which is then use this environment and algorithm to learn it and reinforcements stronger. And we know that robots want to accomplish the mission so they will react to the best action in the next based on the environment change.