We empirically evaluate several state-of-the-art methods for constructing ensembles of classifiers with stacking and show that they perform (at best) comparably to selecting the best classifier from the ensemble by cross validation. We then propose a new method for stacking, that uses multi-response model trees at the meta-level, and show that it outperforms existing stacking approaches, as wel...

We empirically evaluate several state-of-theart methods for constructing ensembles of heterogeneous classifiers with stacking and show that they perform (at best) comparably to selecting the best classifier from the ensemble by cross validation. We then propose a new method for stacking, that uses multi-response model trees at the meta-level, and show that it clearly outperforms existing stacki...

The structure of 2,3,6,7,10,11-hexahydroxytriphenylene (hhtp) methanol monosolvate, C18H12O6·CH3OH, has triclinic symmetry (space group P-1). The compound has a three-dimensional layered network structure formed by intermolecular hydrogen bonding. Structure analysis with Hirshfeld surfaces is shown to be a sensitive method for comparing π-stacking effects in the five known solvates of hhtp. The...

We extend our earlier work on X-ray source stacking in the deep XMM-Newton observation of the Lockman Hole, to the 2 Ms Chandra Deep Field North and the 1 Ms Chandra Deep Field South. The XMM-Newton work showed the resolved fraction of the X-ray background to be ∼ 80-100 per cent at . 2 keV but this decreased to only ∼ 50 per cent above ∼ 8 keV. The CDF-N and CDF-S probe deeper, and are able to...

Supertagging was recently proposed to provide syntactic features for statistical dependency parsing, contrary to its traditional use as a disambiguation step. We conduct a broad range of controlled experiments to compare this specific application of supertagging with another method for providing syntactic features, namely stacking. We find that in this context supertagging is a form of stacking...

We show how high-level scene properties can be inferred from classiication of low-level image features, speciically for the indoor-outdoor scene retrieval problem. We systematically studied the features: (1) his-tograms in the Ohta color space (2) multiresolution, simultaneous autoregressive model parameters (3) coef-cients of a shift-invariant DCT. We demonstrate that performance is improved b...

We show how high level scene properties can be in ferred from classi cation of low level image features speci cally for the indoor outdoor scene retrieval prob lem We systematically studied the features his tograms in the Ohta color space multiresolution si multaneous autoregressive model parameters coef cients of a shift invariant DCT We demonstrate that performance is improved by computing fe...

We have developed a new multicolor fluorescent sensing system to detect multiple analytes in one pot. This design is based on the noncovalent assembly of dye-labeled aptamer with single-walled carbon nanotubes (SWNTs) by π-stacking between the nucleotide bases and the SWNTs sidewalls. In the presence of the targets, the aptamer-target binding separates the assembly of dye-labeled aptamers and S...

The molecular complex lumiflavin-2-aminobenzoic acid monohydrate (C13H12N4O2.C7H7NO2.H2O) crystallizes from from aqueous solution as red triclinic prisms. The space group is P1 with cell dimensions a = 9.660 A, b = 14.866 A, c = 7.045 A, alpha = 95.44 degrees , beta = 95.86 degrees, and gamma = 105.66 degrees . The crystal structure was solved by direct methods and refined by block-diagonal lea...