Article
Environmental Studies
Naroa Coretti Sanchez, Luis Alonso Pastor, Kent Larson
Summary: This study evaluates the environmental impact of autonomous shared bicycles compared to current station-based and dockless systems and finds that autonomy can reduce the environmental impact, highlighting the importance of replacing traditional modes of transportation.
TRANSPORTATION RESEARCH PART D-TRANSPORT AND ENVIRONMENT
(2022)
Article
Energy & Fuels
Jacek Oskarbski, Krystian Birr, Karol Zarski
Summary: This paper introduces a four-stage macroscopic model of bicycle traffic for the city of Gdynia, which takes into account modal shift and the influence of cycling infrastructure characteristics. The model allows for the evaluation of bicycle traffic distribution and demand based on the choice of cycling as an alternative mode of transport.
Article
Computer Science, Information Systems
Dana Kaziyeva, Martin Loidl, Gudrun Wallentin
Summary: This study proposes an agent-based modeling approach to simulate daily bicycle traffic flows in the Salzburg region of Austria. The simulation results demonstrate distinct spatio-temporal patterns of bicycle traffic at high resolution levels. Validation with reference data shows a high correlation between simulated and observed bicycle traffic, emphasizing the importance of input and validation data quality for predictive power.
ISPRS INTERNATIONAL JOURNAL OF GEO-INFORMATION
(2021)
Article
Engineering, Mechanical
Juanjuan Wang, Zishuo Yan, Lili Gui, Kun Xu, Yueheng Lan
Summary: Nonlinear dynamics is a rapidly developing field that studies the spatial and temporal evolution in various disciplines. This article presents a globally valid local approximation method for reconstructing the vector fields of nonlinear systems with unknown parameters or partial observations, based on the invariance of the system's evolution equation. The method demonstrates exceptional robustness and accuracy by considering the interference of noise with nonlinearity to the leading order.
NONLINEAR DYNAMICS
(2022)
Article
Computer Science, Artificial Intelligence
Igor Ilic, Berk Gorgulu, Mucahit Cevik, Mustafa Gokce Baydogan
Summary: Time series forecasting involves developing a model based on past observations to predict future events. The explainable boosted linear regression (EBLR) algorithm enhances predictions by explaining errors and incorporating nonlinear features. It provides interpretable results and high predictive accuracy, making it a promising method for time series forecasting.
PATTERN RECOGNITION
(2021)
Article
Computer Science, Artificial Intelligence
Chang Wei Tan, Christoph Bergmeir, Francois Petitjean, Geoffrey I. Webb
Summary: This paper studies time series extrinsic regression (TSER) and finds that the Rocket TSC algorithm achieves the highest accuracy when adapted for regression. More research is needed to improve the accuracy of ML models in this field, with good prospects for further advancements beyond straightforward baselines.
DATA MINING AND KNOWLEDGE DISCOVERY
(2021)
Article
Green & Sustainable Science & Technology
Chuang Gao, Jiabin Yu, Xiaoguang Zhao, Haibao Wang, Zhiyong Liu, Yaodong Gu
Summary: This study examined the impact of built environment elements on the leisure-time physical activity and walking levels of older people in Ningbo. The results showed significant associations between certain elements of the built environment and the activity levels of older individuals, with variations observed between different sex groups.
Article
Geography, Physical
Qiang Zhou, Zhe Zhu, George Xian, Congcong Li
Summary: Harmonic analysis of time series is important for revealing seasonal land surface dynamics using remote sensing information, but frequency selection can be difficult. The Harmonic Adaptive Penalty Operator (HAPO) is a novel regression method that addresses this issue.
ISPRS JOURNAL OF PHOTOGRAMMETRY AND REMOTE SENSING
(2022)
Article
Mathematics, Applied
Muhammad Amin, Saima Afzal, Muhammad Nauman Akram, Abdisalam Hassan Muse, Ahlam H. Tolba, Tahani A. Abushal
Summary: In data analysis, choosing the right regression model and detecting outliers are crucial for obtaining reliable results. This study proposes new methods for outlier detection in gamma regression (GR) by using adjusted and standardized Pearson residuals. The comparison of existing and proposed methods through simulation and real-life data demonstrates that the adjusted Pearson residual based approach performs better in detecting outliers.
Article
Economics
Xiaohong Chen, Zhijie Xiao, Bo Wang
Summary: This paper studies copula-based time series models that capture both nonstationary and nonlinear patterns in economic and financial time series data. It introduces a procedure that removes nonstationarity through filtration and captures nonlinear temporal dependence using a flexible Markov copula. Two estimators of the copula dependence parameters are proposed and their limiting distributions are analyzed. The results show that the semiparametric copula estimator using filtered data has the same limiting distribution as that without nonstationary filtration, while the parametric copula estimator depends on the nonstationary filtration and may have a non-normal distribution. The robust properties of the semiparametric copula estimators extend to models with misspecified copulas and facilitate statistical inferences in the presence of nonstationarity.
JOURNAL OF ECONOMETRICS
(2022)
Review
Ecology
Hannah S. Wauchope, Tatsuya Amano, Jonas Geldmann, Alison Johnston, Benno Simmons, William J. Sutherland, Julia P. G. Jones
Summary: Human impact on the environment is growing, with increasing strategies to conserve biodiversity. However, there is still a lack of understanding on how interventions affect ecological and conservation outcomes, especially in the analysis of time series data. A standardized framework is needed to robustly assess the impacts of interventions on ecological time series.
TRENDS IN ECOLOGY & EVOLUTION
(2021)
Article
Geosciences, Multidisciplinary
Malgorzata Winska
Summary: This study aims to assess the hydrological effects of polar motion calculated from different combinations of geophysical excitations. The comparison between the observed geodetic excitation function and the atmospheric and oceanic excitation functions reveals differences in estimating hydrological effects. The complex geophysical models used in this study still have uncertainties in the process descriptions, parametrization, and forcing.
FRONTIERS IN EARTH SCIENCE
(2023)
Article
Computer Science, Information Systems
Vitor de Castro Silva, Bruno Bogaz Zarpelao, Eric Medvet, Sylvio Barbon Jr
Summary: A wide range of Machine Learning algorithms can model time series, but they often struggle with characteristics such as repeating patterns and seasonal variations. Current approaches to time series segmentation need improvement and tend to neglect explainability. In this study, we proposed the eXplainable Time Series Tree (XTSTree), which divides a time series into a binary tree based on change detectors. The XTSTree provides a more comprehensive pattern explanation while automatically identifying different time series patterns.
Article
Economics
Phyllis Wan, Richard A. Davis
Summary: Goodness-of-fit tests are commonly used in statistical modeling frameworks to assess the quality of estimated residuals. In the context of time series models, the whiteness of residuals is evaluated using the sample autocorrelation function. This paper applies the auto-distance covariance function (ADCV) to examine the serial dependence of estimated residuals. The limit behavior of the test statistic based on ADCV is derived for a broad class of time series models, considering the adjustment for dependence caused by parameter estimation.
JOURNAL OF ECONOMETRICS
(2022)
Article
Multidisciplinary Sciences
Oleg Uzhga-Rebrov, Peter Grabusts
Summary: The goal of this article is to forecast migration flows in Latvia, particularly focusing on the asymmetric nature of emigration and immigration in the country. Fuzzy time series forecasting methods are employed to predict migration flows in Latvia, as statistical data on migration are often inaccurate. Three different methods are used and a comparative analysis of the results is provided. Generalized forecasts of the expected net migration flow in the future are presented.