[ci] added CODEOWNERS (fixes #2194)

[jameslamb]

See #2194 for details. Please suggest if I've misinterpreted https://github.com/microsoft/LightGBM/blob/76170788c4bd6239334408f88cf51183ec791fa2/docs/FAQ.rst#critical...I did my best to make this file consistent with the conventions I've observed in PRs in this project. It's very likely I have made some mistakes.

My hope is that using CODEOWNERS will turn something being done by convention into something being done by automation. It will, I hope, also improve the contributor experience.

[StrikerRUS]

@jameslamb Thanks for working on it! I like automatization :-)
Please see some comments below.

[jameslamb]

Addressed comments! Please review at your earliest convenience.

[StrikerRUS]

Thanks @jameslamb !
I'm OK with my assignments.

[jameslamb]

rebased to master. If it builds tonight, I'm going to merge this. Thanks to everyone for the reviews!

[jameslamb]

😬 seems one task in azure pipelines has failed. Could someone with sufficient permissions re-run it? It's very very unlikely it could have been caused by this PR.

[StrikerRUS]

😬 seems one task in azure pipelines has failed.

Seems that it was something connected to apt-get, maybe network issue.

Processing triggers for libc-bin (2.19-0ubuntu6.14) ...

bzip2: Compressed file ends unexpectedly;
	perhaps it is corrupted?  *Possible* reason follows.
bzip2: Inappropriate ioctl for device
	Input file = (stdin), output file = (stdout)

It is possible that the compressed file(s) have become corrupted.
You can use the -tvv option to test integrity of such files.

You can use the `bzip2recover' program to attempt to recover
data from undamaged sections of corrupted files.

tar: Child returned status 2
tar: Error is not recoverable: exiting now
sh: 0: Can't open AMD-APP-SDK*.sh

In addition, our master has failed too: https://lightgbm-ci.visualstudio.com/lightgbm-ci/_build/results?buildId=2380.

I have already spotted random failures in GPU task for several times and reported them in #1003. It seems that these bugs occure only in test_cegb_scaling_equalities from time to time.

[jameslamb]

Rebased to master. If this PR builds, I'll merge it tonight!