This time we will introduce the second major machine learning problem. It's called unsupervised learning.
Supervised learning has been introduced before. Looking back at the data set at that time, as shown in the chart, each piece of data in this data set has been marked as negative or positive, that is, benign or malignant tumor. Therefore, for each piece of data in supervised learning, we already know clearly whether the correct answer corresponding to the training set is benign or malignant.
In unsupervised learning, we already know the data. It looks a little different, different from the appearance of supervised learning data, that is, in unsupervised learning there is no label or the same label or no label. So we know the data set, but we don’t know what to do with it, nor are we told what each data point is. Don’t know anything else, it’s just a data set. Can you find some structure in the data? For the data set, unsupervised learning can determine that the data has two different clusters. This is one, that is another, they are different. Yes, unsupervised learning algorithms may separate this data into two different clusters. So it is called clustering algorithm. It turns out that it can be used in many places.
An example of clustering application is in Google News. If you've never seen it before, you can check it out at this URL news.google.com. Google News collects a lot of news content from the Internet every day. It then groups these news into related news. So what Google News does is search a lot of news events and automatically cluster them together. Therefore, these news events are all on the same topic, so they are displayed together.
It turns out that clustering algorithms and unsupervised learning algorithms are also used in many other problems.
Among them is the understanding and application of genetics. a DNA