5 ms·
What kind of experiments have you guys done that combine symbolic and statistical/ML methods? It sounds like an area ripe for research
by voldacar 8y ago
What kind of experiments have you guys done that combine symbolic and statistical/ML methods? It sounds like an area ripe for research
- nextos 8y agoI've built Bayesian non-parametric methods that performed inference on certain formulae, FOPL subsets or even Turing-complete programs. IMHO, it's a very exciting field that will bloom midterm.
- thelastbender12 8y agoThis sounds fun. Do you mind giving a longer description?
- nextos 8y agoI can't say much due to contract. But I can point you to a good source to get started: https://v1.probmods.org/learning-as-conditional-inference.html#example-inferring-an-arithmetic-expression https://v1.probmods.org/learning-as-conditional-inference.ht...
- brundolf 8y agoI know we use ML to "grease the wheels" of inference; i.e., Cyc gains an intuition about what kinds of paths of reasoning to follow when searching for conclusions. I don't know of any higher-level hybridization experiments; I think we only have one ML person on staff and mostly our commercial efforts focus on accentuating what we can do that ML can't, so we haven't had the chance to do many projects where we combine the two as equals.
- hitekker 8y agoHmmm. > I (and my company) believe in a hybrid approach > I don't know of any higher-level hybridization experiments That contradiction and the admission that Cyc only has "one ML person on staff" signals to me, an outsider, that the belief of parity between Machine Learning and "Symbolic" might be predicated more on faith than on reason.
- _bxg1 8y agoI would say "more on theory than on empirical evidence". It's entirely reasonable; the way your eye "thinks" is entirely different from how your higher cognition "thinks", but you need both. If you want something more concrete, here's a recent experiment done by MIT in this realm: https://news.mit.edu/2019/teaching-machines-to-reason-about-what-they-see-0402 https://news.mit.edu/2019/teaching-machines-to-reason-about-... We aren't an ML shop ourselves; we don't claim to be. Given that we have around 100 people, we focus on what we have that's special instead of trying to compete in an overcrowded market. The idea of hybrid AI is something we see as a future part for us to play in the bigger picture of machine intelligence.
- _0ffh 8y agoWow, I should have read further ahead in the comments, before dumping my first thoughts [1] as a standalone. How do you interface between the distinct parts of your machinery? Do you use deeper level neural network representations/activities as symbol embeddings? [1] https://news.ycombinator.com/item?id=19717680 https://news.ycombinator.com/item?id=19717680
- brundolf 8y agoWe're beginning to run up against what I may not be allowed to talk about :) But I will affirm that Cyc is fundamentally symbolics-based. We don't position ourselves as anti-ML, because it's seriously good at a certain subset of things, but Cyc would still be fully-functional without any ML in the picture at all.
- brundolf 8y agoTo clarify the above: "Cyc gains an intuition about what kinds of paths of reasoning to follow when searching for conclusions" The possible paths come purely from symbolics. But that creates a massive tree of possibilities to explore, so ML is used simply to prioritize among those subtrees.
- Iv 8y agoBasically you are learning the heuristic? Do you have any public information on that? That something I have always wanted to work on and really think it could be a shortcut to AGI...
- dhash 8y agohttp://fastsmt.ethz.ch/ http://fastsmt.ethz.ch/