New system allows computers to teach by example

MIT researchers have developed a new system that bridges the ways that computers and humans process information to enable better decision-making.
Computers are good at identifying patterns in huge data sets while humans are good at inferring patterns from just a few examples.
The new system bridges these two ways of processing information, so that humans and computers can collaborate to make better decisions.
The system learns to make judgments by crunching data but distills what it learns into simple examples.
In experiments, human subjects using the system were more than 20 per cent better at classification tasks than those using a similar system based on existing algorithms.

"In this work, we were looking at whether we could augment a machine-learning technique so that it supported people in performing recognition-primed decision-making," said Julie Shah, an assistant professor of aeronautics and astronautics at Massachusetts Institute of Technology and a co-author on the new paper.

"That's the type of decision-making people do when they make tactical decisions - like in fire crews or field operations.
"When they're presented with a new scenario, they don't do search the way machines do. They try to match their current scenario with examples from their previous experience, and then they think, 'OK, that worked in a previous scenario,' and they adapt it to the new scenario," Shah said.

Shah and her colleagues were trying to augment a type of machine learning known as "unsupervised".
In supervised machine learning, a computer is fed a slew of training data that's been labelled by humans and tries to find correlations - say, those visual features that occur most frequently in images labelled 'car'.
In unsupervised machine learning, on the other hand, the computer simply looks for commonalities in unstructured data. The result is a set of data clusters whose members are in some way related, but it may not be obvious how.
The MIT researchers made two major modifications to the type of algorithm commonly used in unsupervised learning.

The first is that the clustering was based not only on data items' shared features, but also on their similarity to some representative example, which the researchers dubbed a 'prototype'.
The other is that rather than simply ranking shared features according to importance, the way a topic-modelling algorithm might, the new algorithm tries to winnow the list of features down to a representative set, which the researchers dubbed a 'subspace'.