Article
Computer Science, Artificial Intelligence
Zhigang Sun, Guotao Wang, Pengfei Li, Hui Wang, Min Zhang, Xiaowen Liang
Summary: In this paper, an improved random forest algorithm based on the classification accuracy and correlation measurement of decision trees is proposed. The algorithm retains decision trees with better classification effects and reduces the correlations between decision trees to improve the performance of the random forest. Experimental results demonstrate that the proposed improved random forest achieves higher average classification accuracy and outperforms traditional random forests in terms of G-means and other metrics.
EXPERT SYSTEMS WITH APPLICATIONS
(2024)
Article
Psychology, Multidisciplinary
Paula Carolina Ciampaglia Nardi, Evandro Marcos Saidel Ribeiro, Jose Lino Oliveira Bueno, Ishani Aggarwal
Summary: This study analyzed the importance of cognitive and financial factors in profit forecasting accuracy by analysts using data from publicly traded Brazilian companies in 2019. Results indicated that optimism, anchoring bias, and commonality had different relationships with accuracy among cognitive biases, while volatility, indebtedness, and profitability were the most important financial factors affecting analyst accuracy. The Random Forest models showed higher explanatory power, enlightening the influence of cognitive and financial aspects on analyst accuracy.
FRONTIERS IN PSYCHOLOGY
(2022)
Article
Environmental Sciences
Ibrahim Fayad, Nicolas Baghdadi, Frederic Frappart
Summary: This study validates the water level estimates from GEDI against in situ gauge station records and analyzes the factors affecting the accuracy. Results show that using viewing angle as a filtering criterion is the best compromise between retained shot count and water level estimation accuracy. V2 provides elevations with a more constant but higher bias compared to V1.
Article
Business
Pratik Arte, Jorma Larimo
Summary: International diversification has a non-linear inverted U-shaped relationship with firm performance. The dual-diversification strategy is detrimental to firm performance, while low/related product diversification benefits firm performance. There is no significant difference in the performance of firms from advanced and emerging economies, but the importance of intangible assets for diversified firms is emphasized.
JOURNAL OF BUSINESS RESEARCH
(2022)
Article
Environmental Sciences
Haruhiko Madarame
Summary: This study investigated changes in shot distribution and accuracy in senior and under-17 international basketball tournaments in the 2010s. Analysis of 1055 games from 20 world-level tournaments revealed significant differences in two-point shot attempts between 2012 and 2016, as well as in three-point shot attempts between 2010, 2012, and 2016. Shot accuracy varied significantly between age categories, with under-17 games showing lower accuracy compared to senior games.
INTERNATIONAL JOURNAL OF ENVIRONMENTAL RESEARCH AND PUBLIC HEALTH
(2021)
Article
Multidisciplinary Sciences
Carmela Bernardo, Lingfei Wang, Francesco Vasca, Yiguang Hong, Guodong Shi, Claudio Altafini
Summary: This paper introduces a dynamic model to describe the achievement of the 2015 Paris Agreement on climate change. The model includes short and long time scales, is able to reach consensus within a similar time horizon, and identifies key leadership roles played by various parties.
Article
Forestry
Wenjing Han, Chang Zhang, Cheng Wang, Luqin Yin
Summary: By clustering the color information of 986 forest images from 40 national forest parks in China using k-means, a forest color palette has been constructed to quantify forest colors and improve their ornamental effect. The study found that forest colors could be divided into eight color families, with green having the highest recognition accuracy for humans and blue-green having the lowest. For future research, the collection pathways for forest images should be expanded, and color information extraction algorithms that incorporate human perception should be selected to improve data availability for forest color studies.
Article
Chemistry, Physical
Wei Yu, Zhaofu Zhang, Xuhao Wan, Jinhao Su, Qingzhong Gui, Hailing Guo, Hong-xia Zhong, John Robertson, Yuzheng Guo
Summary: In this study, an artificial neural network-based potential (NNP) is developed to simulate the Ge3Sb6Te5 material with over 10,000 atoms, and successfully captures the rapid transition of its crystallization process.
CHEMISTRY OF MATERIALS
(2023)
Article
Economics
Virginia Morales Olmos
Summary: Latin America is an important source of global forest values, both in terms of commercial production and forest services. This paper reviews the forest sector data of 19 South and Central American countries plus Mexico from 1990 to 2020, highlighting the distinctions within major producers like Brazil, Chile, and Uruguay, as well as between Latin American countries and other developing countries with large forest endowments. The study emphasizes the implications of these distinctions for the future of the Latin American forest sector.
FOREST POLICY AND ECONOMICS
(2022)
Article
Astronomy & Astrophysics
Xiaohan Wu, Matthew McQuinn, Daniel Eisenstein
Summary: Modern cosmological simulations may face issues like over-ionization and inability to reproduce small-scale fluctuations in the post-reionization ionizing background due to common closure methods. High angular resolution is needed for accurate capturing of post-reionization ionizing background fluctuations, and a strong dependence on the speed of light reduction in simulations suggests a larger ionizing photon mean free path than observational values.
JOURNAL OF COSMOLOGY AND ASTROPARTICLE PHYSICS
(2021)
Article
Psychology, Multidisciplinary
Bing Jia, Zhemin Zhu, Hang Gao
Summary: The cognitive diagnosis model (CDM) can describe the learning progress of students in a specific area, analyzing statistical data from 30,092 students based on PISA 2012 and finding similar learning trajectories in several countries.
FRONTIERS IN PSYCHOLOGY
(2021)
Article
Forestry
Rens G. Brouwer, Pieter A. Zuidema, Fidel Chiriboga-Arroyo, Manuel R. Guariguata, Chris J. Kettle, Francisco Ehrenberg-Azcarate, Julia Quaedvlieg, Mishari R. Garcia Roca, Ronald Corvera-Gomringer, Flor Vargas Quispe, Merel Jansen
Summary: Evaluating Brazil nut restoration projects implemented by smallholders in the Peruvian Amazon, it was found that secondary forests were the most suitable sites for planting Brazil nut trees. Management practices had significant effects on tree growth, survival, and fruit production, emphasizing the importance of post-planting maintenance in increasing the success of forest landscape restoration projects.
FOREST ECOLOGY AND MANAGEMENT
(2021)
Article
Computer Science, Artificial Intelligence
Mamoru Mimura
Summary: There has been a significant increase in malware attacks and malicious traffic. Various machine learning-based detection models have been developed, but their evaluation methods and datasets differ, making it difficult to compare their performances accurately. This study proposes a new metric for evaluating accuracy degradation caused by increasing the benign sample size in binary classification. Using the FFRI dataset, the classification accuracy was evaluated with extracted strings from malware, and it was found that increasing the benign sample size resulted in a decrease in the F1 score.
EXPERT SYSTEMS WITH APPLICATIONS
(2023)
Article
Environmental Sciences
Marguerite C. Pelletier, Michael Charpentier
Summary: The study used M-AMBI to assess the health of estuaries and coasts in the US and found that factors such as water and sediment quality and benthic invertebrates were related to M-AMBI. The study identified variables that were strongly related to M-AMBI at both national and ecoregional scales and highlighted the importance of water clarity, agriculture, and sediment metal concentrations. The results provide valuable information for environmental protection policies and further research.
MARINE POLLUTION BULLETIN
(2023)
Article
Green & Sustainable Science & Technology
Isamara Sousa Conceicao de Benathar, Bruna Virgilio Almeida, Marcos Rodrigues, Jonathan Benathar de Oliveira Sousa, Marcel Regis Moreira da Costa Machado, Angelo Augusto Ebling, Selma Lopes Goulart
Summary: This study aimed to determine the extraction potential of Copaiferaoleoresin in the Carajas National Forest by mapping trees and formulating a logistic regression model. It was found that factors such as tree diameter, dense crown, intact trunk, and presence of damage or bifurcation increased the chances of oleoresin extraction. Sustainable management of forest resources and establishing rational extraction routes and cycles are key contributions of the model.
ENVIRONMENT DEVELOPMENT AND SUSTAINABILITY
(2021)