November 16, 2016
ModiFace-Invented AI Algorithm Taught by Humans Learns Beyond Its Training
A new machine learning training method enables
neural networks to learn directly from human-defined rules, opening new possibilities
for artificial intelligence in fields from medical diagnostics to self-driving cars
A team of ModiFace and University of Toronto scientists have designed an algorithm that learns directly from human instructions,
rather than an existing set of examples, and outperformed conventional methods of
training neural networks by 160 per cent. But more surprisingly, our algorithm
also outperformed its own training by nine per cent - it learned to recognize hair
in pictures with greater reliability than that enabled by the training, marking a
significant leap forward for artificial intelligence.
Our algorithm learned to correctly classify difficult, borderline cases
distinguishing the texture of hair versus the texture of the background.
What we saw was like a teacher instructing a child, and the child learning beyond what the teacher taught her initially.
Humans teach neural networks - computer networks that learn dynamically -
by providing a set of labeled data and asking the neural network to make decisions
based on the samples it's seen. For example, you could train a neural network to
identify sky in a photograph by showing it hundreds of pictures with the sky labeled.
This algorithm is different: it learns directly from human trainers. With this model,
called heuristic training, humans provide direct instructions that are used to pre-classify training samples rather than a set of fixed examples. instructions rather
than a set of examples. Trainers program the algorithm with guidelines such as 'Sky
is likely to be varying shades of blue', and 'Pixels near the top of the image are more
likely to be sky than pixels at the bottom'.
Our work is published today in the online version of the journal IEEE Transactions on Neural Networks
and Learning Systems, with the same version appearing in an upcoming print version of the journal.
This heuristic training approach holds considerable promise for addressing one of
the biggest challenges for neural networks: making correct classifications of
previously unknown or unlabeled data. This is crucial for applying machine learning
to new situations, such as correctly identifying cancerous tissues for medical
diagnostics, or classifying all the objects surrounding and approaching a self-driving