search

tapios / risk-networks

Code for risk networks: a blend of compartmental models, graphs, data assimilation and semi-supervised learning
Other
2 stars 2 forks source link

WIP: learning parameters #180

Closed jinlong83 closed 3 years ago

jinlong83 commented 3 years ago

This PR has the following updates:

  1. enabled the learning of parameters;
  2. implemented batching for the learning of transmission rates;
  3. implemented localization for the learning of transmission rates;
  4. added weighted-averaging-based localization for the learning of transmission rate;
  5. add post-processing tools to analyze the learned parameters (to be done by Jinlong).
odunbar commented 3 years ago

One thing which will still need to be sorted in this PR (If one requires batching) is Issue #169 . (I've described it in detail there.) We need to batch the parameters. It's a little fiddly as parameters are held in a flat numpy array.

One solution would be 2 arrays, one of size [1 x n_glob_params] for the global parameters (to go in every batch) and one of size [n_states x n_nodal_params] for the nodal parameters. Then the batching we can just concatenate the former with the slices [nodes_in_batch , : ] of the latter for the relevant batch.

jinlong83 commented 3 years ago

One thing which will still need to be sorted in this PR (If one requires batching) is Issue #169 . (I've described it in detail there.) We need to batch the parameters. It's a little fiddly as parameters are held in a flat numpy array.

One solution would be 2 arrays, one of size [1 x n_glob_params] for the global parameters (to go in every batch) and one of size [n_states x n_nodal_params] for the nodal parameters. Then the batching we can just concatenate the former with the slices [nodes_in_batch , : ] of the latter for the relevant batch.

Yes you're right, it turns out that learning transition rates leads to computational bottleneck, and the batching for transition rates learning has been implemented in this PR.

jinlong83 commented 3 years ago

This PR was outdated and has been closed. For more recent developments of learning parameters, please refer to PR #189.