4 ms·
>> It seems to me that if the programmer understands the algorithm, then the algorithm is limited by the programmer! Well, a learning algorithm is always going
by YeGoblynQueenne 6y ago
>> It seems to me that if the programmer understands the algorithm, then the
algorithm is limited by the programmer!
Well, a learning algorithm is always going to be limited by something. In AI
literature any such something is grouped under "inductive bias" and there is
no machine learning algorithm that doesn't incorporate some kind of inductive
bias. Bayesian learners have their priors, distance learners have their
disance functions, Support Vector Machines in particular have their kernels,
and of course neural networks have their intricate architectures painstakingly
hand-engineered and fine-tuned to a particular domain, or even a specific
dataset [1]. Indeed it is probably impossible to have learning without
inductive bias [2] [3].
The opinion in the short piece above is representative of a current trend in
machine learning, of taking "the human out of the loop", which loosely
translates in trying to learn everything end-to-end, only from examples, while
pretending that no attempt is made at any point to guide the learner to find
a consistent hypothesis that explains the examples.
Unfotunately, in practicce, this ideal remains a fantasy. All the progress
achieved with deep learning in the last few years would not have been possible
without the discovery (by chance or concerted effort) of good biases that are
conducive to learning in specific domains, e.g. convolutional layers for image
recognition, or Long-Short Term Memory cells for sequence learning, etc.
And where did these good biases come from? Why, from human programmers. Humans
ourselves most likely come equipped with very strong, very useful biases.
We've used those biases and our ability to generalise to come up with powerful
abstractions, such as the laws of physics or mathematics. It took us literally
thousands of years to amass this fortune of knowledge (possibly millions,
during our evolution).
Why would we not use those finely-honed biases of ours and all the knowledge
we've collected to kickstart a new form of intelligence? After all, when we
want fire, we don't sit around waiting for thunder to strike a tree, anymore.
We can start fires on our own. Because of our intelligence- because to build
upon prior knowledge is intelligent.
__________________
[1] See for example the Neural Network Zoo, a collection of neural net
architectures:
https://www.asimovinstitute.org/neural-network-zoo/ https://www.asimovinstitute.org/neural-network-zoo/
Or most of the deep learning literature published in the last few years, where
every little architectural tweak is presented as a major breakthrough.
[2] Famously argued by Tom Mitchell in "The Need for Biases in Learning
Genearlizations":
https://www.semanticscholar.org/paper/The-Need-for-Biases-in-Learning-Generalizations-Mitchell/6cf35ec34efa592f83e3a1b748aea14957fc784a?p2df https://www.semanticscholar.org/paper/The-Need-for-Biases-in...
[3] But also see the discussion between Yan Le Cunn and Christopher Manning:
https://www.youtube.com/watch?v=fKk9KhGRBdI https://www.youtube.com/watch?v=fKk9KhGRBdI
- visarga 6y ago> Why would we not use those finely honed biases of ours and all the knowledge we've collected to kickstart a new form of intelligence? Because priors are hard to embed in a model. For example, CNNs are great for translation invariance, but rotation and scaling don't come out of the box. Why don't they simply add the rotation invariance to the model? Because it's hard to express. Also, human priors are limited. If AlphaGo was to be limited to human priors it would never have surpassed us. The best approach so far is to make a network as free from priors as possible (like the transformers) and let it learn from the data, in essence let it rediscover the convolution or other efficient operations from mountains of data.
- eru 6y ago> If AlphaGo was to be limited to human priors it would never have surpassed us. AlphaGo did surpass humans before AlphaGo Zero.
- visarga 6y agoYes, on self play, but not the supervised model. The supervised model was decent but nowhere near super human. It was used as a starting point for the self play model.
- eru 6y agoEven purely supervised model plus alpha-beta search or Monte Carlo Tree Search has the capability to surpass humans. See eg also how classic chess engines work. (Of course, you could describe alpha-beta search as some form of self-play. But that's probably going to far.)
- YeGoblynQueenne 6y ago>> Because priors are hard to embed in a model. That's a limitation of neural networks, where inductive biases are very difficult to represent. Other approaches don't have any such problem, e.g. all the other approaches I listed above have well-defined, clean and tidy representations for inductive bias. As to AlphaGo, this may come as a shock, but AlphaGo was not the first system to surpass humans in anything. It was the first system to surpass humans _in Go_ but e.g. the first computer system to beat a human grandmaster in chess was Deep Blue [1], in 1997, the first computer system to win a world championship against human players was the checkers (draughts) player Chinook [2], in 1990, the first computer system to outperform humans in medical diagnosis was MYCIN in the 1970's and so on. All those were systems that used strong inductive biases. And of course, AlphaGo itself was limited by human priors- e.g. piece moves, checkerboard dimensions and structure were hard-coded into its architecture and a search for moves was performed by MCTS. In any case, the lack of good enough priors is not a reason to not use any priors- it's a reason to look for better priors. Edit: Has any end-to-end approach rediscovered an entire state-of-the-art architecture, like CNNs or LSTMs? ___________________ [1] https://en.wikipedia.org/wiki/Deep_Blue_(chess_computer) https://en.wikipedia.org/wiki/Deep_Blue_(chess_computer) [2] https://en.wikipedia.org/wiki/Chinook_(computer_program) https://en.wikipedia.org/wiki/Chinook_(computer_program) [3] https://en.wikipedia.org/wiki/Mycin https://en.wikipedia.org/wiki/Mycin