Gaussian Mixture Models (GMMs) are finite mixture models. They are suitable for affine transformation-based probabilistic analysis, such as DC probabilistic power flow.
-
Sparse matrix support: Enhance the library with support for sparse matrices to improve computational efficiency in weak correlation scenarios.
-
More APIs: Expand the library by adding additional APIs to address diverse use cases.
The PDF functions of GMMs are vectorized using the code provided by Gregory Gundersen.
The fast approximation of CDF function of GMMs is from the approxcdf package.
[1] J. T. Flåm, “The Linear Model under Gaussian Mixture Inputs: Selected Problems in Communications,” Doctoral thesis, Norges teknisk-naturvitenskapelige universitet, Fakultet for informasjonsteknologi, matematikk og elektroteknikk, Institutt for elektronikk og telekommunikasjon, 2013. [Online]. Available: https://ntnuopen.ntnu.no/ntnu-xmlui/handle/11250/2370728
[2] Z. Wang, C. Shen, F. Liu, and F. Gao, “Analytical Expressions for Joint Distributions in Probabilistic Load Flow,” IEEE Transactions on Power Systems, vol. 32, no. 3, pp. 2473–2474, May 2017, doi: 10.1109/TPWRS.2016.2612881.