3주차 수요일

neural network 가 image에 optimal한가?

whole slide image-> batch를 사용해서 structure image

tabel data vs. image data

image has "order"

so neural network is not optimal for image 

 

we use CNN

image changes to matrix

https://www.baeldung.com/cs/cnn-feature-map  What Is the Purpose of a Feature Map in a Convolutional Neural Network

---

support vector machine

probelm of linear line : To separate the two classes of data points, there are many possible hyperplanes that could be chosen.

 

Maximizing the margin distance provides some reinforcement so that future data points can be classified with more confidence.

Using these support vectors, we maximize the margin of the classifier.

 

Cost Function and Gradient Updates

In the SVM algorithm, we are looking to maximize the margin between the data points and the hyperplane. The loss function that helps maximize the margin is hinge loss.

Hard and Soft Margin Classification

Forcing rigid margins can result in a model that performs perfectly in the training set, but is possibly over-fit / less generalizable when applied to a new dataset. Identifying the best settings for ‘cost’ is probably related to the specific data set you are working with.

NonLinear SVM Classifier

One approach to handling nonlinear datasets is to add more features.

Luckily we have Kernel Trick, the Kernel Trick is a technique in machine learning to avoid some intensive computation in some algorithms, which makes some computation goes from infeasible too feasible.

---

implement

 

 

- industry

- researcher

- professor

---

 

 

optimization