@misc{HempelKoseskaNikoloskietal.2017,
  author    = {Hempel, Sabrina and Koseska, Aneta and Nikoloski, Zoran and Kurths, J{\"u}rgen},
  title     = {Unraveling gene regulatory networks from time-resolved gene expression data},
  url       = {http://nbn-resolving.de/urn:nbn:de:kobv:517-opus4-400924},
  pages     = {26},
  year      = {2017},
  abstract  = {Background: Inferring regulatory interactions between genes from transcriptomics time-resolved data, yielding reverse engineered gene regulatory networks, is of paramount importance to systems biology and bioinformatics studies. Accurate methods to address this problem can ultimately provide a deeper insight into the complexity, behavior, and functions of the underlying biological systems. However, the large number of interacting genes coupled with short and often noisy time-resolved read-outs of the system renders the reverse engineering a challenging task. Therefore, the development and assessment of methods which are computationally efficient, robust against noise, applicable to short time series data, and preferably capable of reconstructing the directionality of the regulatory interactions remains a pressing research problem with valuable applications. Results: Here we perform the largest systematic analysis of a set of similarity measures and scoring schemes within the scope of the relevance network approach which are commonly used for gene regulatory network reconstruction from time series data. In addition, we define and analyze several novel measures and schemes which are particularly suitable for short transcriptomics time series. We also compare the considered 21 measures and 6 scoring schemes according to their ability to correctly reconstruct such networks from short time series data by calculating summary statistics based on the corresponding specificity and sensitivity. Our results demonstrate that rank and symbol based measures have the highest performance in inferring regulatory interactions. In addition, the proposed scoring scheme by asymmetric weighting has shown to be valuable in reducing the number of false positive interactions. On the other hand, Granger causality as well as information-theoretic measures, frequently used in inference of regulatory networks, show low performance on the short time series analyzed in this study. Conclusions: Our study is intended to serve as a guide for choosing a particular combination of similarity measures and scoring schemes suitable for reconstruction of gene regulatory networks from short time series data. We show that further improvement of algorithms for reverse engineering can be obtained if one considers measures that are rooted in the study of symbolic dynamics or ranks, in contrast to the application of common similarity measures which do not consider the temporal character of the employed data. Moreover, we establish that the asymmetric weighting scoring scheme together with symbol based measures (for low noise level) and rank based measures (for high noise level) are the most suitable choices.},
  language  = {en}
}
@article{ArslanBastiaanseFelser2015,
  author    = {Arslan, Se{\c{c}}kin and Bastiaanse, Roelien and Felser, Claudia},
  title     = {Looking at the evidence in visual world: eye-movements reveal how bilingual and monolingual Turkish speakers process grammatical evidentiality},
  series = {Frontiers in psychology},
  volume    = {6},
  journal   = {Frontiers in psychology},
  publisher = {Frontiers Research Foundation},
  address   = {Lausanne},
  issn      = {1664-1078},
  doi       = {10.3389/fpsyg.2015.01387},
  pages     = {13},
  year      = {2015},
  abstract  = {This study presents pioneering data on how adult early bilinguals (heritage speakers) and late bilingual speakers of Turkish and German process grammatical evidentiality in a visual world setting in comparison to monolingual speakers of Turkish. Turkish marks evidentiality, the linguistic reference to information source, through inflectional affixes signaling either direct (-DI) or indirect (-mls) evidentiality. We conducted an eyetracking-during-listening experiment where participants were given access to visual 'evidence' supporting the use of either a direct or indirect evidential form. The behavioral results indicate that the monolingual Turkish speakers comprehended direct and indirect evidential scenarios equally well. In contrast, both late and early bilinguals were less accurate and slower to respond to direct than to indirect evidentials. The behavioral results were also reflected in the proportions of looks data. That is, both late and early bilinguals fixated less frequently on the target picture in the direct than in the indirect evidential condition while the monolinguals showed no difference between these conditions. Taken together, our results indicate reduced sensitivity to the semantic and pragmatic function of direct evidential forms in both late and early bilingual speakers, suggesting a simplification of the Turkish evidentiality system in Turkish heritage grammars. We discuss our findings with regard to theories of incomplete acquisition and first language attrition.},
  language  = {en}
}
@misc{ArslanBastiaanseFelser2015,
  author    = {Arslan, Se{\c{c}}kin and Bastiaanse, Roelien and Felser, Claudia},
  title     = {Looking at the evidence in visual world},
  series = {Postprints der Universit{\"a}t Potsdam : Humanwissenschaftliche Reihe},
  journal   = {Postprints der Universit{\"a}t Potsdam : Humanwissenschaftliche Reihe},
  number    = {408},
  url       = {http://nbn-resolving.de/urn:nbn:de:kobv:517-opus4-406307},
  pages     = {13},
  year      = {2015},
  abstract  = {This study presents pioneering data on how adult early bilinguals (heritage speakers) and late bilingual speakers of Turkish and German process grammatical evidentiality in a visual world setting in comparison to monolingual speakers of Turkish. Turkish marks evidentiality, the linguistic reference to information source, through inflectional affixes signaling either direct (-DI) or indirect (-mls) evidentiality. We conducted an eyetracking-during-listening experiment where participants were given access to visual 'evidence' supporting the use of either a direct or indirect evidential form. The behavioral results indicate that the monolingual Turkish speakers comprehended direct and indirect evidential scenarios equally well. In contrast, both late and early bilinguals were less accurate and slower to respond to direct than to indirect evidentials. The behavioral results were also reflected in the proportions of looks data. That is, both late and early bilinguals fixated less frequently on the target picture in the direct than in the indirect evidential condition while the monolinguals showed no difference between these conditions. Taken together, our results indicate reduced sensitivity to the semantic and pragmatic function of direct evidential forms in both late and early bilingual speakers, suggesting a simplification of the Turkish evidentiality system in Turkish heritage grammars. We discuss our findings with regard to theories of incomplete acquisition and first language attrition.},
  language  = {en}
}
@phdthesis{Messi2019,
  author    = {Messi, Hugues Urbain Patrick},
  title     = {Les sources du savoir - l'expression de l'inf{\´e}rence en Fran{\c{c}}ais},
  doi       = {10.25932/publishup-46961},
  url       = {http://nbn-resolving.de/urn:nbn:de:kobv:517-opus4-469612},
  school      = {Universit{\"a}t Potsdam},
  pages     = {iii, 291},
  year      = {2019},
  abstract  = {1. Unter Mediativit{\"a}t verstehen wir in dieser Dissertation die sprachliche Markierung der Informationsquelle. Ein Sprecher, der einen Sachverhalt vermittelt, hat die M{\"o}glichkeit durch sprachliche Mittel ausdr{\"u}cklich zu markieren, wie er die {\"u}bermittelte Information bekommen hat. Um diese Informationsquelle sprachlich zu deuten, werden im Franz{\"o}sischen unter anderem einige Verben als mediative Marker (MM) verwendet. 2. Die untersuchten Elemente croire, imaginer, paraitre, penser, savoir, sembler, supposer, trouver sind „mediatiave Verben". Jedes der untersuchten Verben weist besondere semantische und pragmatische Eigenschaften auf, die immer mit dem Ausdruck der Wissensquelle verbunden sind. Es handelt sich also um kognitive Verben (KV), die eine sprachliche Markierung der Informationsquelle vornehmen. Nach ihrem Verhalten in solchen Kontexten erf{\"u}llen sie die Funktion der „mediatiaven Markierung". 3. Die epistemische Modalit{\"a}t ist der Meditivit{\"a}t untergeordnet. Die Erscheinungsform der Modalit{\"a}t (Modalit{\"a}tstyp) bestimmt die St{\"a}rke der epistemischen Modalit{\"a}t. Keines der analysierten Verben dr{\"u}ckt lediglich eine epistemische Leseart aus. Die Dichotomie zwischen der mediativen und epistemischen Modalit{\"a}t besteht darin, dass die erste die Wissensquelle ausdr{\"u}ckt und die zweite ausschließlich die Einstellung des Sprechers gegen{\"u}ber dem Wahrheitsgrad der {\"A}ußerung widerspiegelt. 4. F{\"u}r alle Konstruktionen der Form [V/{\o}P] oder [V, P] ist P die Matrix des Satzes Unsere Ergebnisse zeigen, dass - obwohl diese Konstituenten verschiedene Stellen besetzen k{\"o}nnen - sie dennoch ihre Funktionen als Matrix behalten, indem sie die Propositionen, auf die sie sich beziehen, unter ihrer Rektion behalten. 5. Die Konstruktion [V/{\o}P] und [V, P] stehen in freien Variation Da sich der Wechsel in einem vergleichbaren Kontext vollzieht, und da es in gleicher Umgebung eine freie Substitution gibt, handelt es sich bei den beiden Vorkommen [V/{\o}P] und [V, P] um syntaktische Varianten. 6. Der Konditional-Gebrauch dient haupts{\"a}chlich dazu, die Inferenztypen zu unterscheiden und gleichzeitig die zugrundeliegende Polyphonie zu verdeutlichen. Der Gebrauch des Konditionals dr{\"u}ckt aus, dass es sich nicht um eine zuverl{\"a}ssig zutreffende {\"A}ußerung handelt. Der Ausdruck von Zweifeln kann im Franz{\"o}sischen unter Verwendung spezifischer grammatischer Mittel erfolgen. Zu diesen geh{\"o}rt der Konditional zum Ausdruck der Mitigation (des Zweifels, der Reserviertheit usw.) und der Polyphonie.},
  language  = {fr}
}
@article{VasishthGelman2021,
  author    = {Vasishth, Shravan and Gelman, Andrew},
  title     = {How to embrace variation and accept uncertainty in linguistic and psycholinguistic data analysis},
  series = {Linguistics : an interdisciplinary journal of the language sciences},
  volume    = {59},
  journal   = {Linguistics : an interdisciplinary journal of the language sciences},
  number    = {5},
  publisher = {De Gruyter Mouton},
  address   = {Berlin},
  issn      = {0024-3949},
  doi       = {10.1515/ling-2019-0051},
  pages     = {1311 -- 1342},
  year      = {2021},
  abstract  = {The use of statistical inference in linguistics and related areas like psychology typically involves a binary decision: either reject or accept some null hypothesis using statistical significance testing. When statistical power is low, this frequentist data-analytic approach breaks down: null results are uninformative, and effect size estimates associated with significant results are overestimated. Using an example from psycholinguistics, several alternative approaches are demonstrated for reporting inconsistencies between the data and a theoretical prediction. The key here is to focus on committing to a falsifiable prediction, on quantifying uncertainty statistically, and learning to accept the fact that - in almost all practical data analysis situations - we can only draw uncertain conclusions from data, regardless of whether we manage to obtain statistical significance or not. A focus on uncertainty quantification is likely to lead to fewer excessively bold claims that, on closer investigation, may turn out to be not supported by the data.},
  language  = {en}
}
@misc{SchmidtHesseAttingeretal.2020,
  author    = {Schmidt, Lennart and Heße, Falk and Attinger, Sabine and Kumar, Rohini},
  title     = {Challenges in applying machine learning models for hydrological inference: a case study for flooding events across Germany},
  series = {Postprints der Universit{\"a}t Potsdam : Mathematisch-Naturwissenschaftliche Reihe},
  journal   = {Postprints der Universit{\"a}t Potsdam : Mathematisch-Naturwissenschaftliche Reihe},
  number    = {5},
  issn      = {1866-8372},
  doi       = {10.25932/publishup-52384},
  url       = {http://nbn-resolving.de/urn:nbn:de:kobv:517-opus4-523843},
  pages     = {12},
  year      = {2020},
  abstract  = {Machine learning (ML) algorithms are being increasingly used in Earth and Environmental modeling studies owing to the ever-increasing availability of diverse data sets and computational resources as well as advancement in ML algorithms. Despite advances in their predictive accuracy, the usefulness of ML algorithms for inference remains elusive. In this study, we employ two popular ML algorithms, artificial neural networks and random forest, to analyze a large data set of flood events across Germany with the goals to analyze their predictive accuracy and their usability to provide insights to hydrologic system functioning. The results of the ML algorithms are contrasted against a parametric approach based on multiple linear regression. For analysis, we employ a model-agnostic framework named Permuted Feature Importance to derive the influence of models' predictors. This allows us to compare the results of different algorithms for the first time in the context of hydrology. Our main findings are that (1) the ML models achieve higher prediction accuracy than linear regression, (2) the results reflect basic hydrological principles, but (3) further inference is hindered by the heterogeneity of results across algorithms. Thus, we conclude that the problem of equifinality as known from classical hydrological modeling also exists for ML and severely hampers its potential for inference. To account for the observed problems, we propose that when employing ML for inference, this should be made by using multiple algorithms and multiple methods, of which the latter should be embedded in a cross-validation routine.},
  language  = {en}
}
@article{SchmidtHesseAttingeretal.2020,
  author    = {Schmidt, Lennart and Heße, Falk and Attinger, Sabine and Kumar, Rohini},
  title     = {Challenges in applying machine learning models for hydrological inference: a case study for flooding events across Germany},
  series = {Water Resources Research},
  volume    = {56},
  journal   = {Water Resources Research},
  number    = {5},
  publisher = {John Wiley \& Sons, Inc.},
  address   = {New Jersey},
  pages     = {10},
  year      = {2020},
  abstract  = {Machine learning (ML) algorithms are being increasingly used in Earth and Environmental modeling studies owing to the ever-increasing availability of diverse data sets and computational resources as well as advancement in ML algorithms. Despite advances in their predictive accuracy, the usefulness of ML algorithms for inference remains elusive. In this study, we employ two popular ML algorithms, artificial neural networks and random forest, to analyze a large data set of flood events across Germany with the goals to analyze their predictive accuracy and their usability to provide insights to hydrologic system functioning. The results of the ML algorithms are contrasted against a parametric approach based on multiple linear regression. For analysis, we employ a model-agnostic framework named Permuted Feature Importance to derive the influence of models' predictors. This allows us to compare the results of different algorithms for the first time in the context of hydrology. Our main findings are that (1) the ML models achieve higher prediction accuracy than linear regression, (2) the results reflect basic hydrological principles, but (3) further inference is hindered by the heterogeneity of results across algorithms. Thus, we conclude that the problem of equifinality as known from classical hydrological modeling also exists for ML and severely hampers its potential for inference. To account for the observed problems, we propose that when employing ML for inference, this should be made by using multiple algorithms and multiple methods, of which the latter should be embedded in a cross-validation routine.},
  language  = {en}
}
@article{SchmidtHesseAttingeretal.2020,
  author    = {Schmidt, Lennart and Hesse, Falk and Attinger, Sabine and Kumar, Rohini},
  title     = {Challenges in applying machine learning models for hydrological inference},
  series = {Water resources research},
  volume    = {56},
  journal   = {Water resources research},
  number    = {5},
  publisher = {American Geophysical Union},
  address   = {Washington},
  issn      = {0043-1397},
  doi       = {10.1029/2019WR025924},
  pages     = {10},
  year      = {2020},
  abstract  = {Machine learning (ML) algorithms are being increasingly used in Earth and Environmental modeling studies owing to the ever-increasing availability of diverse data sets and computational resources as well as advancement in ML algorithms. Despite advances in their predictive accuracy, the usefulness of ML algorithms for inference remains elusive. In this study, we employ two popular ML algorithms, artificial neural networks and random forest, to analyze a large data set of flood events across Germany with the goals to analyze their predictive accuracy and their usability to provide insights to hydrologic system functioning. The results of the ML algorithms are contrasted against a parametric approach based on multiple linear regression. For analysis, we employ a model-agnostic framework named Permuted Feature Importance to derive the influence of models' predictors. This allows us to compare the results of different algorithms for the first time in the context of hydrology. Our main findings are that (1) the ML models achieve higher prediction accuracy than linear regression, (2) the results reflect basic hydrological principles, but (3) further inference is hindered by the heterogeneity of results across algorithms. Thus, we conclude that the problem of equifinality as known from classical hydrological modeling also exists for ML and severely hampers its potential for inference. To account for the observed problems, we propose that when employing ML for inference, this should be made by using multiple algorithms and multiple methods, of which the latter should be embedded in a cross-validation routine.},
  language  = {en}
}