Refine
Year of publication
- 2024 (121)
- 2023 (1118)
- 2022 (1974)
- 2021 (2210)
- 2020 (2722)
- 2019 (2575)
- 2018 (2811)
- 2017 (2517)
- 2016 (2352)
- 2015 (2166)
- 2014 (1884)
- 2013 (2103)
- 2012 (1993)
- 2011 (2047)
- 2010 (1445)
- 2009 (1835)
- 2008 (1363)
- 2007 (1396)
- 2006 (1804)
- 2005 (1956)
- 2004 (2021)
- 2003 (1553)
- 2002 (1354)
- 2001 (1425)
- 2000 (1686)
- 1999 (1852)
- 1998 (1690)
- 1997 (1541)
- 1996 (1557)
- 1995 (1473)
- 1994 (1031)
- 1993 (405)
- 1992 (255)
- 1991 (169)
- 1990 (16)
- 1989 (28)
- 1988 (22)
- 1987 (23)
- 1986 (16)
- 1985 (12)
- 1984 (15)
- 1983 (31)
- 1982 (10)
- 1981 (9)
- 1980 (10)
- 1979 (15)
- 1978 (9)
- 1977 (12)
- 1976 (7)
- 1975 (3)
- 1974 (2)
- 1973 (2)
- 1972 (2)
- 1971 (2)
- 1970 (1)
- 1958 (1)
Document Type
- Article (35104)
- Doctoral Thesis (6459)
- Monograph/Edited Volume (5509)
- Postprint (3218)
- Review (2280)
- Part of a Book (984)
- Other (869)
- Preprint (566)
- Conference Proceeding (526)
- Part of Periodical (450)
Language
- English (29921)
- German (25777)
- Spanish (362)
- French (329)
- Italian (113)
- Russian (112)
- Multiple languages (65)
- Hebrew (36)
- Portuguese (25)
- Polish (24)
Keywords
- Germany (196)
- climate change (171)
- Deutschland (137)
- European Union (78)
- Sprachtherapie (77)
- machine learning (75)
- diffusion (74)
- Patholinguistik (73)
- patholinguistics (73)
- morphology (72)
Institute
- Institut für Biochemie und Biologie (5348)
- Institut für Physik und Astronomie (5316)
- Institut für Geowissenschaften (3537)
- Institut für Chemie (3437)
- Wirtschaftswissenschaften (2634)
- Historisches Institut (2482)
- Department Psychologie (2310)
- Institut für Mathematik (2128)
- Institut für Romanistik (2105)
- Sozialwissenschaften (1882)
Klimaschutzklagen
(2023)
The Paris Agreement enshrines the triad of mitigation, adaptation and resilience. In this connection, climate protection lawsuits are increasingly being filed. The question for the courts is whether or not they are functionally overburdened in view of the dimension of climate change and transformation. They will have to find a path between non-functional activism and misplaced restraint. Adapted from the usual introductory question for civil law cases, «Who wants what out of what from whom?», an overview of the subjects of dispute will be provided. Subsequently, the challenges of the courts of lower instance and the Federal Constitutional Court with regard to the separation of powers and the principle of democracy are presented. Particular attention is paid to the climate decision of the Federal Constitutional Court of March 24th, 2021.
We address classification problems for which the training instances are governed by an input distribution that is allowed to differ arbitrarily from the test distribution-problems also referred to as classification under covariate shift. We derive a solution that is purely discriminative: neither training nor test distribution are modeled explicitly. The problem of learning under covariate shift can be written as an integrated optimization problem. Instantiating the general optimization problem leads to a kernel logistic regression and an exponential model classifier for covariate shift. The optimization problem is convex under certain conditions; our findings also clarify the relationship to the known kernel mean matching procedure. We report on experiments on problems of spam filtering, text classification, and landmine detection.
We address classification problems for which the training instances are governed by an input distribution that is allowed to differ arbitrarily from the test distribution-problems also referred to as classification under covariate shift. We derive a solution that is purely discriminative: neither training nor test distribution are modeled explicitly. The problem of learning under covariate shift can be written as an integrated optimization problem. Instantiating the general optimization problem leads to a kernel logistic regression and an exponential model classifier for covariate shift. The optimization problem is convex under certain conditions; our findings also clarify the relationship to the known kernel mean matching procedure. We report on experiments on problems of spam filtering, text classification, and landmine detection.
One of the main problems in machine learning is to train a predictive model from training data and to make predictions on test data. Most predictive models are constructed under the assumption that the training data is governed by the exact same distribution which the model will later be exposed to. In practice, control over the data collection process is often imperfect. A typical scenario is when labels are collected by questionnaires and one does not have access to the test population. For example, parts of the test population are underrepresented in the survey, out of reach, or do not return the questionnaire. In many applications training data from the test distribution are scarce because they are difficult to obtain or very expensive. Data from auxiliary sources drawn from similar distributions are often cheaply available. This thesis centers around learning under differing training and test distributions and covers several problem settings with different assumptions on the relationship between training and test distributions-including multi-task learning and learning under covariate shift and sample selection bias. Several new models are derived that directly characterize the divergence between training and test distributions, without the intermediate step of estimating training and test distributions separately. The integral part of these models are rescaling weights that match the rescaled or resampled training distribution to the test distribution. Integrated models are studied where only one optimization problem needs to be solved for learning under differing distributions. With a two-step approximation to the integrated models almost any supervised learning algorithm can be adopted to biased training data. In case studies on spam filtering, HIV therapy screening, targeted advertising, and other applications the performance of the new models is compared to state-of-the-art reference methods.
Ciliate epibionts associated with crustacean zooplankton are widespread in aquatic systems, but their ecological roles are little known. We studied the occurrence of ciliate epibionts on crustacean zooplankton in nine German lakes with different limnological features during the summer of 2011. We also measured the detachment and re-attachment rates of the ciliates, changes in their motility, and the feeding rates of attached vs. detached ciliate epibionts. Epibionts were found in all lakes sampled except an acidic lake with large humic inputs. Epibiont prevalence was as high as 80.96% on the cladoceran Daphnia cucullata, 67.17% on the cladoceran Diaphanosoma brachyurum, and 46.67% on the calanoid copepod Eudiaptomus gracilis. Both cladoceran groups typically had less than 10 epibionts per individual, while the epibiont load on E. gracilis ranged from 1 to >30 epibionts per individual. After the death of the zooplankton host, the peritrich ciliate epibiont Epistylis sp. detached in an exponential fashion with a half-life of 5 min, and 98% detached within 30 min, leaving behind the stalks used for attachment. Immediately after detachment, the ciliates were immotile, but 62% became motile within 60 min. When a new host was present, only 27% reattached after 120 min. The average measured ingestion rate and clearance rate of Epistylis were 11,745 bacteria ciliate(-1) h(-1) and 24.33 mu l ciliate(-1) h(-1), respectively. Despite their high feeding rates, relatively low epibiont abundances were observed in the field, which suggests either diversion of energy to stalk formation, high metabolic loss by the epibionts, or high mortality among the epibiont populations.
Zooplankton support distinct bacterial communities in high concentrations relative to the surrounding water, but little is known about how the compositions and functionalities of these bacterial communities change through time in relation to environmental conditions. We conducted a year-long field study of bacterial communities associated with common zooplankton groups as well as free-living bacterial communities in the York River, a tributary of Chesapeake Bay. Bacterial community genetic fingerprints and their carbon substrate usage were examined by denaturing gradient gel electrophoresis (DGGE) of amplified 16S rDNA and by Biolog EcoPlates, respectively. Zooplankton-associated communities were genetically distinct from free-living bacterial communities but utilized a similar array of carbon substrates. On average, bacteria associated with different zooplankton groups were genetically more similar to each other within each month (65.4% similarity) than to bacterial communities of the same zooplankton group from different months (28 to 30% similarity), which suggests the importance of ambient environmental conditions in shaping resident zooplankton-associated bacterial communities. Monthly changes in carbon substrate utilization were less variable for zooplankton-associated bacteria than for free-living bacteria, suggesting that the zooplankton microhabitat is more stable than the surrounding water and supports specific bacterial groups in the otherwise unfavorable conditions in the water column.
This cumulative habilitation thesis presents new work on the systematics, paleoecology, and evolution of antelopes and other large mammals, focusing mainly on the late Miocene to Pleistocene terrestrial fossil record of Africa and Arabia. The studies included here range from descriptions of new species to broad-scale analyses of diversification and community evolution in large mammals over millions of years. A uniting theme is the evolution, across both temporal and spatial scales, of the environments and faunas that characterize modern African savannas today. One conclusion of this work is that macroevolutionary changes in large mammals are best characterized at regional (subcontinental to continental) and long-term temporal scales. General views of evolution developed on records that are too restricted in spatial and temporal extent are likely to ascribe too much influence to local or short-lived events. While this distinction in the scale of analysis and interpretation may seem trivial, it is challenging to implement given the geographically and temporally uneven nature of the fossil record, and the difficulties of synthesizing spatially and temporally dispersed datasets. This work attempts to do just that, bringing together primary fossil discoveries from eastern Africa to Arabia, from the Miocene to the Pleistocene, and across a wide range of (mainly large mammal) taxa. The end result is support for hypotheses stressing the impact of both climatic and biotic factors on long-term faunal change, and a more geographically integrated view of evolution in the African fossil record.
Compressions and extensions
(1998)
75 WR stars and 164 RSGs are identified in a single WFC3 pointing of our M101 survey. We find that within it's large star-forming complex NGC 5462 WR stars are preferentially located in the core whilst RSGs are found in the halo, suggesting two bursts of star-formation. A review of our WR candidates reveals that only ∼30% are detected in the archival broad-band ACS imaging whilst only ∼50% are associated with HII regions.
The occurrence of two individuals of Zu cristatus at 2 m depth in coastal waters of the Gulf of Castellamare (Tyrrhenian Sea, Italy) together with records of this rare pan-Oceanic mesopelagic species is reported. Analyses of two mitochondrial genes (12 s and 16 s; 936 bp) revealed a 2.6% sequence divergence between Mediterranean and Pacific (Japanese) samples of the species.
By combining morphology, ecology, biology, and biogeography with the available molecular (sequence variation of the entire mitochondrial cytochrome b gene; cyt-b) and karyology data, the taxonomy of several species of the Rutilus complex inhabiting southern Europe is revised. Rutilus stoumboudae, new species, is described from Lake Volvi, Greece. It differs from Rutilus rutilus in possessing more total GR and less branched rays in both dorsal and anal fins and in its placement in the cyt-b based phylogeny of the genus. The resurrected genus Leucos Heckel, 1843 (type species Leucos aula, Bonaparte, 1841), which according to molecular data diverged from Rutilus more than 5 million years ago, during the Messinian salinity crisis, includes five species of small size, without spinous tubercles on scales and head in reproductive males, pharyngeal teeth formula 5-5, and all show a preference for still waters. Leucos aula is the Italian species endemic in the Padany-Venetian district: L. basak is widespread in Croatia, Albania, Montenegro and former Yugoslav Republic of Macedonia (FYROM); L. albus, recently described from Lake Skadar, Montenegro, is also found in rivers Moraca and Zeta (Montenegro). L. albus differs from L. basak, its closest relative, in having more scales on the LL and less anal-fin rays; L. panosi is endemic to the western-Greece district, and L. ylikiensis is endemic to lakes Yliki and Paralimni in eastern Greece (introduced in Lake Volvi). Among the nominal species examined, Rutilus karamani, R. ohridanus, R. prespensis and R. prespensis vukovici are all junior synonyms of Leucos basak. Rutilus vegariticus is definitively regarded as junior synonym for R. rutilus. Sarmarutilus n.gen. is a monotypic genus, with Sarmarutilus rubilio as the type species. According to phylogenetic data, Sarmarutilus rubilio is basal to a cluster of species that includes Leucos basak, L. albus, L. aula, L. panosi and L. ylikiensis. Sarmarutilus possibly evolved in pre-Messinian time, in the Lago Mare, entered the Mediterranean area during the Messinian Lago Mare phase of the Mediterranean Sea and survived only in the Tuscany-Latium district. This genus differs from Leucos in having large pearl organs on the central part of head and body scales in mature males and for the habitat preference, being a riverine-adapted species. It differs from Rutilus in pharyngeal teeth formula (5-5 in Sarmarutilus and 6-5 in Rutilus), size (small in Sarmarutilus and large in Rutilus) and for the preferential habitat (riverine vs. still water). Finally, lectotypes for Leucos basak, Leucos aula, and Sarmarutilus rubilio are designated.
Von der Ausnahmesituation der Intellektuellen : ostdeutsche Diskurse und die Weimarer Republik
(1996)