Here is a plot for the next 10 units with 100 simulations.
This is my submission for the Intel OneAPI Code Maven Hackahton conducted by Techgig. The problem statement was to use the OneAPI library to implement a Stock Prediction algorithm that uses Monte Carlo Sampling.
I implemented the paper Modelling the probability of future stock returns- TOBIAS BRODD, ADRIAN DJERF.
I also implemented the bayesian inference method described in this talk here. I tried creating a bayesian inference engine with Hamiltonian Monte Carlo but it didn't work(check this commit to try it for yourself). I also implement Metropolis Hasting for the same but that didn't work either.
To build locally on windows
- Install OneAPI DPC and DPL libraries
- Install CPR and boost-date-time (using vcpkg is most convinient)
To run it just download it from here, generate a API from Rapid API. Put it in a file an use it on the first screen.