Sankarasubramanian Ragunathan

1) Developed a LSTM-SDE model that forecasts the OHLC values of a given asset over a forecast window given the input historical window. The model also incorporates uncertainty in the predictions made thereby making it suitable for formulating trading strategies that mitigate risk.2) Developed "Neural SDEs as GANs" model that generated "fake" trading signals that mimic the actual behavior of the asset movement. This method is beneficial in generating pseudo-asset price data in cases where we do not have acce

Pursuing my PhD in developing higher-order numerical integrators for solving stochastic differential equations. Applications include finance, physics, biology, and so forth.