6 ms·
That is... an unusual use of regression, to say the least. When I see something sufficiently off the wall, I always wonder how the authors happened to think of
by cba9 11y ago
That is... an unusual use of regression, to say the least. When I see something sufficiently off the wall, I always wonder how the authors happened to think of that. It also raises a lot of questions like, how much of physics can be usefully approximated by some random forests? Could you replace most of a physics engine with an appropriate small neural network?
- wisty 11y agoPredictor-correctors is the principal behind a lot of physics engines. I guess they just got a better predictor.
- archgoon 11y agoActually, what they have here is a way of generating predictors from the corrector; which theoretically can be optimized for your domain.
- jimfleming 11y agoCheck out NeuroAnimator[0] from 1997 (also early work from Hinton). It covers some of this using local-spaced hierarchies of neural networks that predict the deltas for the next time-step. [0] http://web.cs.ucla.edu/~dt/papers/siggraph97-sketch/ http://web.cs.ucla.edu/~dt/papers/siggraph97-sketch/
- jasonwatkinspdx 11y agoRandom Forests are part of how the LHC found the higgs, based on my armchair understanding of the slide decks. Measurement, Monte Carlo and Machine Learning form an interesting triangle. My impression of the LHC pipe is they used monte carlo model sims to train up classifiers that would flag the data relevant for distinguishing alternate models.
- ssantic 11y agoNot to hijack the thread, but would you mind posting a link to the LHC slide deck you mentioned?
- jasonwatkinspdx 11y agoI watched the announcement live stream. I'm sure the video and slides are up somewhere. Sorry I don't have it at hand.
- phreeza 11y agoThe major difference there is that LHC data measures processes which are inherently stochastic (because QM), so using these methods is pretty natural. The OP, however, is applying the same methods to highly complex, but nominally deterministic problems.
- noobermin 11y agoI'm curious about this. A lot of us in the computational world have to deal with able but slow iterative methods for things like fluid dynamics and electromagnetic simulations. I'd be interested to know if this works with MHD, and how well it does.
- phreeza 11y agoIt probably would, but it is harder to get performance guarantees. Seems like a taylor expansion on steroids, but with the taylor expansion you get analytical tractability (to prove correct convergence) along with speed, which you don't get in this case.
- dsfsdfd 11y agoIt occurred to me and I work as a web developer. Just think of things in abstract enough terms and possibilities like this pop out. Having the technical sophistication and time to do the work is another matter.
- dr_zoidberg 11y agoNeural nets are seen as universal function approximators when you study them in detail. While it isn't the usual "neural net news" you see posted online, this is exactly what ANNs are meant to do.
- bmh100 11y agoGiven that ANNs are universal function approximators, it is natural that one would use them to actually generate a simplified model of simulations. What you see here, taking an already accurate model and generating an approximate implementation, is definitely rare but not unexpected. Long-running simulations can have their run-time drastically compressed simply by developing a neural network capable of being parallelized on GPUs, as in this paper. The ANN run time could be further compressed by generating a single-hidden-layer equivalent. Even though they are known to be exponentially larger than multiple hidden layers, the implementation can be valuable in real-time systems where analysis latency is at a premium, e.g. image recognition in machine control systems.