Review
Management
Martin Cousineau, Vedat Verter, Susan A. Murphy, Joelle Pineau
Summary: In the absence of randomized controlled and natural experiments, it is necessary to balance the distributions of covariates in order to obtain an unbiased estimate of a causal effect. Optimization-based methods have shown improvement in balancing covariate distributions and estimating causal effects, but thorough comparisons with other methods are lacking. Operations researchers can contribute their knowledge of optimization to improving causal inference methods.
EUROPEAN JOURNAL OF OPERATIONAL RESEARCH
(2023)
Article
Ecology
Fabien Moustard, Muki Haklay, Jerome Lewis, Alexandra Albert, Marcos Moreu, Rafael Chiaravalloti, Simon Hoyte, Artemis Skarlatidou, Alice Vittoria, Carolina Comandulli, Emmanuel Nyadzi, Michalis Vitos, Julia Altenbuchner, Megan Laws, Raffaella Fryer-Moreira, Daniel Artus
Summary: The Sapelli smartphone application is designed to support communities in engaging in citizen science activities to address local concerns. Through the Extreme Citizen Science methodology based on participatory design, users can work with professional scientists to determine challenges, data collection methods, and analysis techniques to address identified problems.
FRONTIERS IN ECOLOGY AND EVOLUTION
(2021)
Article
Chemistry, Physical
Lee Organick, Bichlien H. Nguyen, Rachel McAmis, Weida D. Chen, A. Xavier Kohll, Siena Dumas Ang, Robert N. Grass, Luis Ceze, Karin Strauss
Summary: Synthetic DNA has emerged as a viable alternative for digital data storage, with different physical properties compared to biological DNA. This study evaluated nine preservation methods for synthetic DNA-encoded data files, finding no practical distribution difference between them. Additionally, a discussion on the trade-offs between stability and density of these preservation methods was provided.
Review
Multidisciplinary Sciences
Farzane Amirmahani, Nasim Ebrahimi, Fatemeh Molaei, Ferdos Faghihkhorasani, Kiarash Jamshidi Goharrizi, Seyede Masoumeh Mirtaghi, Marziyeh Borjian-Boroujeni, Michael R. Hamblin
Summary: Translational medicine involves converting findings from basic scientific studies into clinical research, but progress has been slower than expected, and there has been limited clinical efficacy in return on investment.
ANNALS OF THE NEW YORK ACADEMY OF SCIENCES
(2021)
Article
Geosciences, Multidisciplinary
Liyang Xiong, Sijin Li, Guoan Tang, Josef Strobl
Summary: Terrain is a crucial natural geographic feature that plays a significant role in physical processes. The fields of geomorphometry and terrain analysis have provided abundant topographic data and tools, which have contributed to the understanding of geomorphology, hydrology, soil science, and geographic information systems (GIS). Recent advancements in analysis theory, methods, data acquisition techniques, and analysis platforms have facilitated the interpretation of topographic characteristics and associated mechanisms and processes. Future directions include addressing challenges in data collection and construction, expanding the application of efficient analysis frameworks and tools, and integrating terrain analysis with other disciplines for cross-analysis purposes.
EARTH-SCIENCE REVIEWS
(2022)
Article
Multidisciplinary Sciences
Misuk Kim, Kyu-Baek Hwang
Summary: In the context of imbalanced classification problems, the effectiveness and applicability of oversampling methods have not been rigorously evaluated. In this study, we assessed different combinations of oversampling methods and machine learning classifiers using various datasets. Surprisingly, we found that oversampling had limited impact on classifier performance and could even reduce performance. Additionally, oversampling had a greater adverse effect on AUPRC compared to AUROC. Undersampling performed poorly and oversampling was more effective for linear classifiers. Most importantly, we found that oversampling was not necessary to obtain the optimal classifier for the majority of datasets.
Article
Computer Science, Artificial Intelligence
Ruchika Malhotra, Juhi Jain
Summary: This study aims to develop accurate and effective software defect prediction models. The study finds that imbalanced data and a large number of software metrics degrade the model performance. Correlation feature selection and resampling methods can improve the predictive capability of SDP models.
PEERJ COMPUTER SCIENCE
(2022)
Article
Green & Sustainable Science & Technology
Eric A. Jensen, Kalina Borkiewicz, Jill P. Naiman, Stuart Levy, Jeff Carpenter
Summary: This essay demonstrates the application of evidence-based science communication process in creating scientific data visualizations for public audiences. The effectiveness of visualizing research data can benefit millions of viewers, and involving the intended audience in decision-making is crucial. The Advanced Visualization Lab at the University of Illinois at Urbana-Champaign presents their steps towards evidence-based practice by using audience research in designing cinematic-style data visualizations for scientific documentary films, resulting in effective techniques and better understanding of what works and why.
Article
Engineering, Aerospace
A. Pignalberi, M. Pietrella, M. Pezzopane, J. B. Habarulema
Summary: A new data assimilation procedure has been implemented in the IRI UP method to update F2-peak ionospheric characteristics over the South-African region. Results show that the use of different vTEC calibration methods impacts the IRI UP method, with the Ciraolo method modeling foF2 with higher precision.
ADVANCES IN SPACE RESEARCH
(2021)
Article
Environmental Sciences
Oktawia Lewicka, Mariusz Specht, Andrzej Stateczny, Cezary Specht, Czeslaw Dyrcz, Pawel Dabrowski, Bartosz Szostak, Armin Halicki, Marcin Stateczny, Szymon Widzgowski
Summary: This article explores the measurements and conversion of data from the coastal zone using various methods. The study presents different transformation methods and evaluates their effects on the tombolo measurement campaign in Sopot conducted in 2018. The results suggest that the adjustment calculus method is most effective for plane coordinates, while the method of P.S. Dabrowski et al. is best for height coordinates. It is recommended to develop a new transformation method based on the synthesis of these two existing methods.
Article
Multidisciplinary Sciences
Joyce de Souza Zanirato Maia, Ana Paula Arantes Bueno, Joao Ricardo Sato
Summary: Educational indicators are used to evaluate the quality of the educational system and their impact is often associated with economic and social factors. This study aimed to assess factors related to school performance using a dataset from Brazilian schools, generating different models to predict performance. The non-linear model was found to predict the best performance, and the heterogeneity of variable importance across school cycles and regions could help inform the development of public educational policies.
Review
Chemistry, Multidisciplinary
Hadi Parastar, Roma Tauler
Summary: This Review summarizes the analysis of big (bio)chemical data using multivariate chemometric methods and discusses the challenges faced by analytical research. It emphasizes the potential of chemometric methods in solving BBCD problems in chromatographic, spectroscopic, and hyperspectral imaging measurements, with applications to omics sciences. The review provides insights on addressing analysis of BBCD, obtaining reliable qualitative and quantitative results. It discusses the importance of big data in (bio)chemistry, presents analytical tools for generating BBCD, and the theoretical background and limitations of chemometric methods when applied to BBCD. The review highlights the importance of chemometric methods in analyzing BBCD in different chemical disciplines with examples.
ANGEWANDTE CHEMIE-INTERNATIONAL EDITION
(2022)
Article
Energy & Fuels
Cyrus Ashayeri, Birendra Jha
Summary: This study introduces a novel approach of utilizing data-driven analytics for the evaluation of unconventional oil and gas resources, applying a Transfer Learning framework based on a combination of unsupervised and supervised machine learning techniques. By quantifying the changes in predicted productivity and considering the generalizability of different sub-basin models, the study emphasizes the importance of incorporating domain expertise in petroleum engineering and geology in the process.
JOURNAL OF PETROLEUM SCIENCE AND ENGINEERING
(2021)
Article
History & Philosophy Of Science
Sharon Crasnow
Summary: The passage discusses the reliance of empirical research on democracy on data, and the development and evaluation of measurement models. Using the V-Dem project as an example, the author argues that assessing the reliability and validity of measurement models is challenging, and proposes the concept of "coherence objectivity" where democracy can be understood objectively under theoretical, empirical, and pragmatic constraints.
Article
Computer Science, Artificial Intelligence
Kushal Kanti Ghosh, Shemim Begum, Aritra Sardar, Sukdev Adhikary, Manosij Ghosh, Munish Kumar, Ram Sarkar
Summary: DNA microarray experiments provide information about cell and tissue states, with only a few genes playing a significant role in disease classification. Feature selection algorithms aim to efficiently identify relevant features, with feature ranking techniques assigning importance to features without using learning algorithms. This paper extensively studies 10 popular filter ranking methods and their performance on various microarray datasets using different classifiers. The experiments show that Mutual Information is the most effective method among Entropy based methods, ReliefF is best in the Similarity based methods category, and Chi-square performs well in the Statistics based methods category.
EXPERT SYSTEMS WITH APPLICATIONS
(2021)
