Learning Solution of Differential Equations using LSSVM based model

- Machine learning models can learn the trajectories of a dynamical systems in semi-supervised fashion!

- Only few training data which correspond to the initial and boundary conditions are needed!

- The solution in the rest of domain will be learned!

- Suitable for spatio-temporal models and also high dimensional PDEs.

For those interested in further investigating the connections between Machine Learning and Differential Equations feel free to consult with the following papers:

- S. Mehrkanoon, J. A. K. Suykens, “Learning solutions to partial differential equations using LS-SVM,” Neurocomputing, Vol. 159, pp. 105-116, 2015. [PDF]

- S. Mehrkanoon, J. A. K. Suykens, "LS-SVM approximate solution to linear time varying descriptor systems", Automatica, 48(10), 2502-2511, 2012. [PDF][Slides]

- S. Mehrkanoon, T. Falck, J. A. K. Suykens, "Approximate solutions to ordinary differential equations using least squares support vector machines", IEEE Trans. Neural Netw. Learning Syst, 23(9), 1356-1367, 2012. [PDF]