764 search results - page 74 / 153 » Learning Models for Multi-Source Integration |
COLT
2000
Springer
15 years 4 months ago
2000
Springer
We study model selection strategies based on penalized empirical loss minimization. We point out a tight relationship between error estimation and data-based complexity penalizatio...
GECCO
2006
Springer
15 years 3 months ago
2006
Springer
This paper proposes an adaptive discretization method, called Split-on-Demand (SoD), to enable the probabilistic model building genetic algorithm (PMBGA) to solve optimization pro...
110
click to vote
IROS
2008
IEEE
15 years 6 months ago
2008
IEEE
Abstract— Bayesian filtering is a general framework for recursively estimating the state of a dynamical system. The most common instantiations of Bayes filters are Kalman filt...
101
click to vote
COMPSYSTECH
2007
15 years 3 months ago
2007
: The paper proposes a different approach to data modeling. Analogous to the rejection method, where the misclassifications are removed and manually evaluated, we focus here on dif...
EDM
2008
15 years 1 months ago
2008
This paper presents the development of two related machine-learned models which predict (a) whether a student can answer correctly questions in an ILE without requesting help and (...