Hello there..

Popular methods include support vector regression and random forest regression. They are good, though need some tweaking in parameter settings. For first trial, I suggest to use methods with minimum parameters such as random forest. Later, try to explore techniques with large parameter settings such as ANN, SVR, ...

HTH

Hello and thank you!

If anyone is interested, there is also Cubist software that builds rule-based predictive models.

There are a number of publications where it is used with remote sensing data.

It is a commercial software, but source code for a single-threaded Linux version is available under the GNU General Public License.

Dear Vasily

Please check this article:

A comparative assessment of multi-temporal Landsat 8 and machine learning algorithms for estimating aboveground carbon stock in coppice oak forests

http://www.tandfonline.com/doi/abs/10.1080/01431161.2017.1356488

Regards,