The current lambda calculation procedure calculates lambdas serially, which reduces the overall performance of the algorithm. It would be preferable to have a number of lambdas calculated, up through the number of workers. This is dependent on #7 to synchronize the model state among the models as well as #11 to rework how the data is handled among the workers.
The current lambda calculation procedure calculates lambdas serially, which reduces the overall performance of the algorithm. It would be preferable to have a number of lambdas calculated, up through the number of workers. This is dependent on #7 to synchronize the model state among the models as well as #11 to rework how the data is handled among the workers.