【Python Code】Use the nonlinear support vector machine to classify the identified model parameters
This article is recommended for such people !
  • Person who wants to write program of nonlinear support vector machine
  • A person studying machine learning to solve a classification problem
  • A person performing model parameter identification for research

Hi, This is a research archive article, but please see if you want to see this.
So, what am I doing ?

  • Modeling the EEG signal
  • Identify model parameters by solving the inverse problem
  • Classification by SVM to determine concentration or relax state

I’m doing something like this.
If you want to know more detail, please go to this page.

Actually, I am writing a dissertation, so I can not write specific data on my blog.

But, python program code is managed in my blog.
So, you can use it for free.

I’m showing some unused data.
Red marks are concentrated state and blue marks are relax state.

Classification is possible for data that can be linearly separated.

The left side shows the data used to draw classification lines while the right side shows the data to test weather the line works well.
The accuracy rate is about 87 % (It is given above).

I think that experimental data itself is not very good, so it is necessary to recover experimental data again.

Programming of nonlinear support vector machine

I will write a program for nonlinear support vector machine!

I commended the code, so it might be understandable

SVM parameters

In this code, the kernel method is described by pipeline.
The data of X_train and y_train are used by the following SVM parameters..

Polynominal kernel

For high-dimensional mapping of feature quantities, we use 3-dimensional mapping using a polynomial kernel.

Standard Scalar

We are standardizing pre-processing of data

Standardize to compare other condition data.
That’s all the important.