April 27, 2018
View f492651

refactor almost duplicated code into new function: update_interpolated_losses_in_interval

Jorn Hoofwijk authored on 27/04/2018 19:29:36
View e0c2fd2

make more clear this is a range

Jorn Hoofwijk authored on 27/04/2018 19:25:03
View 4465c93

remove some part of the code that did something with interpolated points

Jorn Hoofwijk authored on 27/04/2018 19:19:08
View 7091113

only update_losses once for each point

Jorn Hoofwijk authored on 27/04/2018 19:14:54
View 0298267

don't change scale change factor

Jorn Hoofwijk authored on 27/04/2018 19:14:34
April 24, 2018
View d3cc7c8

now it also works with runners

Jorn Hoofwijk authored on 24/04/2018 14:59:32
View 1cc1cb1

interpolate loss, and do not interpolate the actual points

Jorn Hoofwijk authored on 24/04/2018 14:34:18
April 19, 2018
View 65ca5f7

Merge branch 'bugfix/termination'

Joseph Weston authored on 19/04/2018 17:48:10
View 73b3628

consistently treat machine-precision sized intervals in Learner1D

Joseph Weston authored on 18/04/2018 17:50:46
View 655fd0e

interpolate the loss instead of the function

Jorn Hoofwijk authored on 19/04/2018 14:33:30
April 18, 2018
View b46bc2b

correct point choosing in the first few steps of the 1DLearner

Joseph Weston authored on 18/04/2018 17:26:29
March 19, 2018
View f9a65ab

plot the Learner1D with margins such that the boundary points are visible

Bas Nijholt authored on 19/03/2018 14:25:57
March 7, 2018
View 0759344

1D: extrapolate the data and do local updates instead of recalculating all interpolated values

Bas Nijholt authored on 07/03/2018 23:09:28
February 19, 2018
View 4965811

make loss functions accessible from learner namespace

Joseph Weston authored on 19/02/2018 18:37:32
View 39864b5

add uniform sampling loss functions

Joseph Weston authored on 19/02/2018 16:11:42