Computational Methods for Jump Processes
From Monday 7th of July 2014 (lunch time) to Wednesday 9th of July (afternoon).
Diﬀusion and jump diﬀusion models are used to describe a large variety of stochastic phenomena in both natural and social sciences from ﬁnance and credit risk to neurobiology and chemistry. However, designing eﬃcient inference and simulation procedures for diﬀusions and jump diﬀusions is a very challenging task. Some elementary properties of these models, such as transition probabilities, are not available in closed form. Consequently, exact simulation of diﬀusion trajectories and diﬀusion bridges is highly nontrivial. Models that account for jumps add another layer of complication.
The aim of this workshop is to highlight the recent advances in the estimation and the simulation of jump diﬀusions. All aspects of inference for stochastic processes will be considered:
- Series approximation based methods.
- Spectral/generator based simulation and estimation methods for jump processes.
- Perfect simulation of jump processes.
- Sequential Monte Carlo methods.
A poster session will take place on Tuesday 8th July. The abstract must contain all relevant points of your presentation and is limited to 300 words. The deadline for submission is Januray 31st.
A few number of fundings for junior researchers (a maximum of 2 years of experience since completion of Ph.D.) will be available to cover registration and accomodation. All applicants must join a CV with their abstract submission.
If you have any questions, please email Sylvain Le Corff:
sylvain dot lecorff at gmail dot com
This workshop is supported by CRiSM.