Offical implementation of Dynamic Ensemble (DyEnsemble) neural decoding algorithm (NeurIPS 2019).
Qi, Y., Liu, B., Wang, Y., & Pan, G. (2019). Dynamic ensemble modeling approach to nonstationary neural decoding in Brain-computer interfaces. In Advances in Neural Information Processing Systems (pp. 6089-6098).
Implemented by Zhuo Chen (Zhejiang University, https://github.com/zhuo34).
Here is an example for how to run DyEnsemble with codes in main.py.
Firstly, create customized transition functions and measurement functions by inheriting DyEnsembleTModel and DyEnsembleMModel. In this example, transition functions and measurement functions are shown as follows.
Secondly, generate simulation data.
Thirdly, create DyEnsemble.
# model = DyEnsemble(x_dim, z_dim, t_model, *m_models)
model = DyEnsemble(x_dim, z_dim, F(x_dim, step), H1(z_dim), H2(z_dim), H3(z_dim))Finally, use DyEnsemble.
# model(x_idle, z, n_particle, alpha, save_pm=True):
model(0, y, 200, alpha[i], save_pm=True)Run main.py by python main.py, then you will get a figure in which each row represents the posterior probability of each mesurement model(function), and different columns represent different alpha.
