MoT - Mixture of trees probabilistic graphical model for video segmentation

We present a novel mixture of trees (MoT) graphical model for video segmentation. Each component in this mixture represents a tree structured temporal linkage between super-pixels from the first to the last frame of a video sequence. Our time-series model explicitly captures the uncertainty in temporal linkage between adjacent frames which improves segmentation accuracy. We provide a variational inference scheme for this model to estimate super-pixel labels and their confidences in nearly realtime. The efficacy of our approach is demonstrated via quantitative comparisons on the challenging SegTrack joint segmentation and tracking dataset [23].

Characterization of aconitine-type alkaloids in the flowers of Aconitum kusnezoffii by electrospray ionization tandem mass spectrometry

The fragmentation mechanism of aconitine-type alkaloids in the flowers of Aconitum kusnezoffii (FAK) was investigated using electrospray ionization tandem mass spectrometry (ESI-MSn) firstly. The analysis of the collision-induced dissociation (CID) spectra of three purified aconitine standards and six previously reported aconitines indicated that the fragmentation of the protonated aconitines at low-energy CID follows a similar pathway. The elimination of a C-8-substituent such as an acetic acid or a fatty acid is the dominant fragmentation mode in MS2. Successive losses of CH3COOH, CH3OH, H2O, BzOH, and CO are the main fragmentation pathways of aconitine-type alkaloids in MS3 spectra. Based on these features, a rapid method for the direct detection and characterization of alkaloids from an ethanolic extract of FAK is described. All the known aconitum alkaloids are detected and a series of lipo-aconitines has been found for the first time in this plant.

Altitudinal differences in the leaf fitness of juvenile and mature alpine spruce trees (Picea crassifolia)

In many plant species, leaf morphology varies with altitude, an effect that has been attributed to temperature. It remains uncertain whether such a trend applies equally to juvenile and mature trees across altitudinal gradients in semi-arid mountain regions. We examined altitude-related differences in a variety of needle characteristics of juvenile (2-m tall) and mature (5-m tall) alpine spruce (Picea crassifolia Kom.) trees growing at altitudes between 2501 and 3450 m in the Qilian Mountains of northwest China. We found that stable carbon isotope composition (delta C-13), area- and mass-based leaf nitrogen concentration (N-a, N-m), number of stomata per gram of nitrogen (St/N), number of stomata per unit leaf mass (St/LM), projected leaf area per 100 needles (LA) and leaf mass per unit area (LMA) varied nonlinearly with altitude for both juvenile and mature trees, with a relationship reversal point at about 3 100 m. Stomatal density (SD) of juvenile trees remained unchanged with altitude, whereas SD and stomatal number per unit length (SNL) of mature spruce initially increased with altitude, but subsequently decreased. Although several measured indices were generally found to be higher in mature trees than in juvenile trees, N-m, leaf carbon concentration (C.), leaf water concentration. (LWC), St/N, LA and St/LM showed inconsistent differences between trees of different ages along the altitudinal gradient. In both juvenile and mature trees, VC correlated significantly with LMA, N-m, N-a, SNL, St/LM and St/N. Stomatal density, LWC and LA were only significantly correlated with delta C-13 in mature trees. These findings suggest that there are distinct ecophysiological differences between the needles of juvenile and mature trees that determine their response to changes in altitude in semi-arid mountainous regions. Variations in the fitness of forests of different ages may have important implications for modeling forest responses to changes in environmental conditions, such as predicted future temperature increases in high attitude areas associated with climate change.

A Constant-Factor Approximation Algorithm for Embedding Unweighted Graphs into Trees

We present a constant-factor approximation algorithm for computing an embedding of the shortest path metric of an unweighted graph into a tree, that minimizes the multiplicative distortion.

Hierarchical Multi-classification with Predictive Clustering Trees in Functional Genomics

Struyf, J., Dzeroski, S. Blockeel, H. and Clare, A. (2005) Hierarchical Multi-classification with Predictive Clustering Trees in Functional Genomics. In proceedings of the EPIA 2005 CMB Workshop

Fuzzy-Rough Feature Significance for Fuzzy Decision Trees

R. Jensen and Q. Shen, 'Fuzzy-Rough Feature Significance for Fuzzy Decision Trees,' in Proceedings of the 2005 UK Workshop on Computational Intelligence, pp. 89-96, 2005.