1 / 10

Efficient Multimodal Bayesian Learning for Broad Knowledge Acquisition

Learn how large-scale Bayesian learning from text, images, and video can enhance knowledge acquisition for AGI through probabilistic models and inference methods. Explore the potential of incorporating built-in and learned components to scale up and improve real-world data analysis.

ullery
Download Presentation

Efficient Multimodal Bayesian Learning for Broad Knowledge Acquisition

An Image/Link below is provided (as is) to download presentation Download Policy: Content on the Website is provided to you AS IS for your information and personal use and may not be sold / licensed / shared on other websites without getting consent from its author. Content is provided to you AS IS for your information and personal use only. Download presentation by click this link. While downloading, if for some reason you are not able to download a presentation, the publisher may have deleted the file from their server. During download, if you can't get a presentation, the file might be deleted by the publisher.

E N D

Presentation Transcript


  1. AGI Through Large-Scale, Multimodal Bayesian Learning Brian Milch MIT March 2, 2008

  2. Need for Broad Knowledge How can I get from Boston to New Haven without a car? Broad and deep world knowledge How many U.S. congress members have PhDs? About how cold is it in this picture? Image (c) ukdave.com

  3. Acquiring Such Knowledge • Proposal: Learn knowledge from large amountsof online text, images, video • Learn by Bayesian belief updating, maintaining probability distribution over: • Models of how world tends to work • Past, current, and future states of the world

  4. Variables in Probability Model Tendencies Language Use Appearances … … … World History Utterance Structure Scene Structure … … … … … … … Linguistic Data Video Data

  5. Data to Learn From • Text? • Lots available; broad coverage • No connection with sensory input • Experience of physical or virtual robot(s)? • Multimodal; get to actively manipulate world • Physical: hard to get broad experience • Virtual: may not generalize to physical world • Multimodal data on the Web • Broad coverage; linguistic and sensory • Disjointed; sometimes not factual; passive + − + − − + −

  6. Built-In Components • Why built-in components? • Children don’t learn from scratch • Why not exploit known algorithms, data structures (rendering, parse trees, …) • Modules for reasoning about: • Space, time, physical objects, shape • Language • Other agents

  7. Learned Components Dirichlet process prior allowing models to grow to explain data [Kemp et al., AAAI 2006] Tendencies Relational probabilistic models [Getoor & Taskar 2007] with initially unknownobject types, predicates, dependencies … World History … … Structures with initially unknown objects, relations and attributes (over time) … … … … …

  8. Algorithms • Probabilistic inference • Markov chain Monte Carlo [Gilks et al. 1996] • Variational methods [Jordan et al. 1999] • Belief propagation [Yedidia et al. 2001] • Hybrids with logical inference [Sang et al. 2005; Poon & Domingos 2006] • Parallelize interpretation of documents, images, videos • Still unclear how to scale up sufficiently

  9. Measures of Progress • Should be able to show steady improvement on real data sets (object recognition, coreference, entailment, …) • Serve as resource for shallower, hand-built systems (replacing Cyc, WordNet) • Spin off challenges for researchers in specialty areas

  10. Conclusions • Potential path to AGI: Bayesian learning on large amounts of multimodal data • Attractive features • Exploits well-understood principles • Learns broad, real-world knowledge • Connected to mainstream AI research

More Related