필터 지우기
필터 지우기

How can I use the Support Vector Machine (SVM) for making classification of trajectories of many objects

조회 수: 1 (최근 30일)
Dear Sir,
I want to make trajectories' classification of many objects by using SVM classifier. I am still new in using SVM classifier. I made check on the internet and I find simple example by using SVM classifier "http://www.mathworks.it/it/help/stats/support-vector-machines-svm.html" but I don't understand in the end if example how the SVM make classification. I would like to know "Can you please support me with some example (links) about SVM applications (clear applications) to guide me after that to make classifications to many different trajectories??
Thank you very much
Mohamed Elwakdy

답변 (1개)

Shashank Prasanna
Shashank Prasanna 2013년 10월 9일
The link you provided has two full examples of classification using SVM. Did you try running those? Do you have specific questions after you ran the examples?
It is really as simple as:
% Step 1: Training
SVMStruct = svmtrain(Training,Group)
% Step 2: Classification
newGroup = svmclassify(SVMStruct,newSample)
newGroup has the classified information for the new data newSample.
Try out the examples in the documentation and then try different problems from datasets here:
  댓글 수: 4
Mohamed Elwakdy
Mohamed Elwakdy 2013년 10월 14일
Dear Shashank,
Thank you very much on your interest...
I would like to know "Do you know any SVM application that used any ship trajectories database??" From another side, I would like to know "Do you any website that I can from it on ship trajectories database over the time ??"
All the ship trajectories that I obtained on it by Google search whether are not over the time or the number of rows or columns (when I make preparation of data) are not equal between the training dataset and checking dataset so I have a problem to use the SVM where the results are not accuracy.
Thank you very much
Have Good Day
Shashank Prasanna
Shashank Prasanna 2013년 10월 14일
Unfortunately I am not aware of another website for your examples. If you can post your code and data we can take a look at it.

댓글을 달려면 로그인하십시오.

Community Treasure Hunt

Find the treasures in MATLAB Central and discover how the community can help you!

Start Hunting!

Translated by