12943 search results - page 163 / 2589 » Predicting When Not to Predict |
125
Voted
ICIG
2009
IEEE
15 years 7 months ago
2009
IEEE
Categorizing multiple objects in images is essentially a structured prediction problem: the label of an object is in general dependent on the labels of other objects in the image....
112
Voted
IDA
2009
Springer
15 years 7 months ago
2009
Springer
Abstract. This work aims to improve an existing time series forecasting algorithm –LBF– by the application of frequent episodes techniques as a complementary step to the model....
93
Voted
ICPR
2008
IEEE
15 years 7 months ago
2008
IEEE
One of the major limitations of HMM-based models is the inability to cope with topology: When applied to a visible observation (VO) sequence, HMM-based techniques have difficulty ...
101
click to vote
ICTAI
2007
IEEE
15 years 7 months ago
2007
IEEE
In this paper we developed an Inductive Logic Programming (ILP) based framework ExOpaque that is able to extract a set of Horn clauses from an arbitrary opaque machine learning mo...
89
Voted
SMC
2007
IEEE
15 years 7 months ago
2007
IEEE
— This paper predicts that gaze and fractal dimension of images may offer some of measurability to when humans look at still or moving images expanding before one’s eyes such a...