top of page
Blog: Blog2
Search
Writer's picturethuytien nguyen

In my defense I was left unsupervised shirt

In my defense I was left unsupervised shirt supervised learning in a natural way almost all examples we see online implement supervised learning. They assume all the data they have belongs to a class Y. The goal is to identify those two clusters in an unsupervised manner. Then with the fictitious data points added, they run a decision tree algorithm to partition the space as shown in Fig B. The reason this works is if there are indeed clusters in our original data, it cannot be uniformly distributed. So by adding the fictitious data points, we can identify the true clusters we need because within those clusters the number of Y points will be more than the fictitious points. These stand for convolutional neural networks. This is a special type of neural network, that is designed for data with spatial structure.

📷

2 views0 comments

コメント


bottom of page