A new numerical method for 1-D backward stochastic differential equations without using conditional expectations
Keyword(s):
AbstractIn this paper, we propose a new numerical method for 1-D backward stochastic differential equations (BSDEs for short) without using conditional expectations. The approximations of the solutions are obtained as solutions of a backward linear system generated by the terminal conditions. Our idea is inspired from the extended Kalman filter to non-linear system models by using a linear approximation around deterministic nominal reference trajectories.
2014 ◽
Vol 15
(3)
◽
pp. 618-646
◽
2006 ◽
Vol 28
(4)
◽
pp. 1563-1581
◽
2002 ◽
Vol 11
(4)
◽
pp. 972-995
◽
2011 ◽
Vol 29
(6)
◽
pp. 1008-1032
◽
2021 ◽
Vol 37
(7)
◽
pp. 1156-1170