[ML] Line search for best learn rate

[tveasey]

This switches to performing an initial line search over the learn rate and reducing the range we search in the main optimisation loop. This does cost us some run time, but we generally get small wins in both the QoR and the stability of QoR as borne out by the sample results below.

Because we can't estimate the runtime upfront I've extended the progress monitor to allow us to reset the expected total number of steps. This will advance progress if it needs to; we have no mechanism for reducing progress if the expected number steps increase. In this case, we simply get stuck at the current progress. I use a margin in the initial estimates to reduce the chance of this so we get smoother progress monitoring.

Sample regression results:

data set	master mean R^2	master sd R^2	patch mean R^2	patch var R^2
boston	0.85	0.011	0.92	0.006
elevators	0.88	0.028	0.89	0.002
stock	0.98	0.004	0.99	0.003
houses	0.68	0.026	0.66	0.016

Sample classification results:

data set	master mean acc	master sd acc	patch mean acc	patch var acc
bank	0.87	0.003	0.88	0.006
census	0.84	0.004	0.84	0.003
clicks	0.66	0.016	0.67	0.009
qsar	0.84	0.007	0.85	0.008

[valeriy42]

Good work altogether! I am wondering if adding mainLoop as prefix or suffix actually improves readability since very often there is no alternative variable, like mainLoopMaximumNumberTrees and initializationLoopMaximumNumberTrees. The exception beeing, totalNumberSteps and mainLoopNumberSteps.

[tveasey]

Thanks for the review @valeriy42. I've now addressed all your comments can you take another look?

[valeriy42]

LGTM. Good work on providing explanations!