Respected madams, I want to how to apply the Support Vector Machines in Wind Forecasting. I need to develop model using the classification algorithm. But i am not getting idea about how to use it. Hence kindly guide me in this regard. It will be helpful for my research.
Dear Sir,
To use any classification algorithm for forecasting (including SVM) first you need to prepare your data correctly. You need input data (X) and output data (Y). Depending on the data you have, your input might be wind direction from different weather stations, atmospheric pressures, humidity etc. All those entries would be input values.The output value would be wind direction in the site of your intrest which corresponds to the collected input data. So one ith point in n dimensional space would have the following shape:
{Xi,Yi} = {{PressureStation1i, WindDirectionStation1,..., PressureStationni, WindDirectionStationni,..., },WindDirectionToBePredictedi}
The data should be normalized by columns (0-1 for example), however when using some good wrapper for SVMs (like Mathematica) there's no need to normalize the data, it is done automatically. The data should look somewhat like in the pdf file I attached.
It is possible that lag-time alanysis might be required when dealing with time series.
Your data is the learning set which is used to train the support vector machine.
Many informations on general use of SVMs can be found on the LibSVM website.
https://www.csie.ntu.edu.tw/~cjlin/libsvm/
Determine the number of classes and label them. Then, select a subset of labeled samples as training samples and a subset of samples or all of them as testing samples. Use the LIBSVM tools for classification with SVM, the link of LIBSVM is: http://www.csie.ntu.edu.tw/~cjlin/libsvm.
The attached files help you to use LIBSVM.
had the question addressed for sirs also I would have answered. However, for now kindly refer to the following link
http://users.ecs.soton.ac.uk/srg/publications/pdf/SVM.pdf
Dear Wojciech Artichowicz, Maryam Imani,Harsh Bhasin Md. Matiur Rahaman thank you very much for your answers.
- Badges
- Science topic
- Similar topics
- Organizational Psychology
- Training