en-de
en-es
en-fr
en-pt
en-sl
en
en-zh
0.25
0.5
0.75
1.25
1.5
1.75
2
Learning from constraints
Published on Oct 03, 20114094 Views
In this talk, I propose a functional framework to understand the emergence of intelligence in agents exposed to examples and knowledge granules. The theory is based on the abstract notion of constrain
Related categories
Chapter list
Learning from Constraints: Bridging perception and symbolic reasoning00:00
Acknowledgments00:33
Merging of two ideas ...01:15
Outline02:30
Environment and constraints02:57
Labeled examples02:58
Diagnosis and prognosis in medicine04:02
Handwritten char discrimination04:32
Text categorization05:14
Optical flow06:23
(Logic) constraints07:11
(Logic) constraints (con’t)08:37
Agents and constraints09:01
Equivalence10:43
Learning from (given) constraints11:35
Ambient space11:56
Semi-norm in Sobolev spaces13:48
Parsimony Principle16:10
Soft-constraints17:43
Representation (hard constraints)18:06
Representation21:39
Representation (soft constraints)21:42
Two remarkable "examples"21:59
Where do kernel machines come from ...23:36
When kernels arise from regularization operators ...24:43
Lagrange multipliers and probability density25:56
Putting things in context by unsupervised data27:16
Collapse of dimensionality29:42
Case studies31:26
Multi-intervals31:56
Box kernels33:07
Changing the regularization parameter34:58
From Gaussian (plain kernel)36:05
Linear (plain kernel)36:18
Asset allocation36:39
Look at the residual ...37:36
Perceptual and logic constraints ...38:23
With supervised examples only / With logic constraints39:25
The effect of forcing logic constraints39:41
Two-stages ...40:06
Conclusions41:15
Do you want to know more?41:20