Improving classification using a Confidence Matrix based on weak classifiers applied to OCR

This paper proposes a new feature representation method based on the construction of a Confidence Matrix (CM). This representation consists of posterior probability values provided by several weak classifiers, each one trained and used in different sets of features from the original sample. The CM allows the final classifier to abstract itself from discovering underlying groups of features. In this work the CM is applied to isolated character image recognition, for which several set of features can be extracted from each sample. Experimentation has shown that the use of CM permits a significant improvement in accuracy in most cases, while the others remain the same. The results were obtained after experimenting with four well-known corpora, using evolved meta-classifiers with the k-Nearest Neighbor rule as a weak classifier and by applying statistical significance tests.

Y-box binding protein 1 (YB-1) relevance in estrogen receptor-positive (ER+) breast cancer

Tese de mestrado, Biologia Molecular e Genética, Universidade de Lisboa, Faculdade de Ciências, 2016

Temporal Control of Leaf Complexity by miRNA-Regulated Licensing of Protein Complexes

The tremendous diversity of leaf shapes has caught the attention of naturalists for centuries. In addition to interspecific and intraspecific differences, leaf morphologies may differ in single plants according to age, a phenomenon known as heteroblasty. In Arabidopsis thaliana, the progression from the juvenile to the adult phase is characterized by increased leaf serration. A similar trend is seen in species with more complex leaves, such as the A. thaliana relative Cardamine hirsuta, in which the number of leaflets per leaf increases with age. Although the genetic changes that led to the overall simpler leaf architecture in A. thaliana are increasingly well understood, less is known about the events underlying age-dependent changes within single plants, in either A. thaliana or C. hirsuta. Here, we describe a conserved miRNA transcription factor regulon responsible for an age-dependent increase in leaf complexity. In early leaves, miR319-targeted TCP transcription factors interfere with the function of miR164-dependent and miR164-independent CUC proteins, preventing the formation of serrations in A. thaliana and of leaflets in C. hirsuta. As plants age, accumulation of miR156-regulated SPLs acts as a timing cue that destabilizes TCP-CUC interactions. The destabilization licenses activation of CUC protein complexes and thereby the gradual increase of leaf complexity in the newly formed organs. These findings point to posttranslational interaction between unrelated miRNA-targeted transcription factors as a core feature of these regulatory circuits.

Organic geochemistry on surface sediments from the Laptev Sea

Sediment samples from the Laptev Sea, taken during the 1993 RV Polarstern expedition ARK IX/4 and the RV Ivan Kireyev expedition TRANSDRIFT I, were investigated for the amount and composition of their organic carbon fractions. Of major interest was the identification of different processes controlling organic carbon deposition (i.e. terrigenous supply vs. surface water productivity). Long-chain unsaturated alkenones derived from prymnesiophytes, and fatty acids derived from diatoms and dinoflagellates, were analysed by means of gas chromatography and mass spectrometry. First results on the distribution of these biomarkers in surface sediments indicate that the surface water productivity signal is well preserved in the sediment data. This is shown by the distribution of the 16:1(n-7) and 20:5(n-3) fatty acids indicative for diatoms, and the excellent correlation with the chlorophyll a concentrations in the surface water masses and the biogenic-opal content and increased hydrogen indices of the sediments. The high concentration of these unsaturated fatty acids in shallow water sediments shows the recent deposition of the organic material. In deep-sea sediments, on the other hand, the concentrations are low. This decreased content is typical for phytoplankton material which has been degraded by microorganisms or autoxidation. In general, the alkenone concentrations are very low, suggesting low production rates by prymnesiophytes. Only at one station from the lower continental margin influenced by the inflow of Atlantic water masses, were some higher amounts of alkenones determined. Long-chain n-alkanes as well as high C/N ratios and low hydrogen indices indicate the importance of (fluvial) supply of terrigenous organic matter.

Macerals in surface sediments of the Laptev Sea (Arctic Ocean)

Surface sediments from the Laptev Sea and adjacent continental slope were studied for their composition of particulate organic matter (OM) by means of maceral analysis. The composition of macerals in sediments gives information about the environment, terrigenous supply from the hinterland, and marine OM. With reference to their biological sources, we distinguish between terrigenous and marine macerals. We found that the particulate OM in the surface sediments of the Laptev Sea is predominantly of terrigenous origin (mean: 78%). However, distinct variations exist when looking in detail. In the shelf area, sediments may contain up to 99% terrigenous OM. Freshwater algae occur directly north of the river mouths, reflecting the strong fluvial influence. Relatively high amounts of marine OM (20-40%) are restricted to the upper continental slope, the Vilkitsky Strait and west of the New Siberian Islands, explained by increased surface-water productivity due to increased fluvial nutrient supply, open-water conditions, and phytoplankton blooms at the ice-edge.

Distribution of clay minerals in surface sediments of the Eurasian Arctic Ocean

Clay-mineral distributions in the Arctic Ocean and the adjacent Eurasian shelf areas are discussed to identify source areas and transport pathways of terrigenous material in the Arctic Ocean. The main clay minerals in Eurasian Arctic Ocean sediments are illite and chlorite. Smectite and kaolinite occur in minor amounts in these sediments, but show strong variations in the shelf areas. These two minerals are therefore reliable in reconstructions of source areas of sediments from the Eurasian Arctic. The Kara Sea and the western part of the Laptev Sea are enriched in smectite, with highest values of up to 70% in the deltas of the Ob and Yenisey rivers. Illite is the dominant clay mineral in all the investigated sediments except for parts of the Kara Sea. The highest concentrations with more than 70% illite occur in the East Siberian Sea and around Svalbard. Chlorite represents the clay mineral with lowest concentration changes in the Eastern Arctic, ranging between 10 and 25%. The main source areas for kaolinite in the Eurasian Arctic are Mesozoic sedimentary rocks on Franz-Josef Land islands. Based on clay-mineral data, transport of the clay fraction via sea ice is of minor importance for the modern sedimentary budget in the Arctic basins.