News for hackers. Stuff that matters.
Re:A tutorial on active learning (2009) [pdf] by jmmcd on Saturday October 28, 2017 @07:04AM (#15574713)
Great tutorial. I see active learning as a halfway house between supervised learning and reinforcement learning, because requesting labels is an action (as in RL), but of a very limited, predefined type.

A lot of problems which we initially model as supervised learning are in reality, in a live situation, more like active learning. Going the whole hog to RL may be unnecessary. But still, bandit-type models may be the right fully general setting.

These topics are very much in the air at the moment with a lot of new interest in bandit models and algorithms.

Re:A tutorial on active learning (2009) [pdf] by mycat on Saturday October 28, 2017 @07:09AM (#15574724)
Are there decent recent implementations in python of these? Or tutorials of some sort?
Re:A tutorial on active learning (2009) [pdf] by charlescearl on Saturday October 28, 2017 @10:39AM (#15575271)
The vowpal wabbit library provides an active learning setup https://github.com/JohnLangford/vowpal_wabbit/wiki/active_le...

I think you launch an active learning server, and the Python application interacts with it.

You may be able to run everything in Python https://github.com/JohnLangford/vowpal_wabbit/tree/master/py....

Here's further description https://github.com/JohnLangford/vowpal_wabbit/wiki/Command-l...