Article
Computer Science, Information Systems
Wan-Lei Zhao, Hui Wang, Peng-Cheng Lin, Chong-Wah Ngo
Summary: This paper addresses the issue of merging k-nearest neighbor (k-NN) graphs in two different scenarios. A symmetric merge algorithm is proposed to combine two approximate k-NN graphs, facilitating large-scale processing. A joint merge algorithm is also proposed to expand an existing k-NN graph with a raw dataset, enabling the incremental construction of a hierarchical approximate k-NN graph.
IEEE TRANSACTIONS ON BIG DATA
(2022)
Article
Computer Science, Hardware & Architecture
Martin Aumueller, Sariel Har-Peled, Sepideh Mahabadi, Rasmus Pagh, Francesco Silvestri
Summary: This paper studies the r-NN problem in similarity search in the context of individual fairness and equal opportunities. The authors propose efficient data structures for the fair NN problem and highlight the inherent unfairness of existing NN data structures through experimental evaluation.
COMMUNICATIONS OF THE ACM
(2022)
Article
Automation & Control Systems
Hongjiao Guan, Long Zhao, Xiangjun Dong, Chuan Chen
Summary: Imbalanced data classification is a challenging problem in many applications. We propose an extended natural neighbor (ENaN) concept without parameter k to improve the quality of generated examples by accurately reflecting the local distribution. ENaN-based SMOTE (ENaNSMOTE) can improve the sample distribution obtained by SMOTE and NaNSMOTE.
ENGINEERING APPLICATIONS OF ARTIFICIAL INTELLIGENCE
(2023)
Article
Computer Science, Artificial Intelligence
Jianping Gou, Liyuan Sun, Lan Du, Hongxing Ma, Taisong Xiong, Weihua Ou, Yongzhao Zhan
Summary: This article proposes a novel representation coefficient-based k-nearest centroid neighbor method (RCKNCN) aiming to improve the classification performance and reduce the sensitivity to the neighborhood size k. The method captures both the proximity and geometry of k-nearest neighbors and learns to differentiate the contribution of each neighbor to the classification of a testing sample. A weighted majority voting algorithm is also proposed under the RCKNCN framework.
EXPERT SYSTEMS WITH APPLICATIONS
(2022)
Article
Computer Science, Information Systems
Wan-Lei Zhao, Hui Wang, Chong-Wah Ngo
Summary: This paper presents a simple yet effective solution for approximate k-nearest neighbor search and graph construction. The solution integrates graph construction and search tasks, and supports dynamic updates on the built graph.
IEEE TRANSACTIONS ON MULTIMEDIA
(2022)
Article
Automation & Control Systems
Amit Kumar Gangwar, Om Prakash Mahela, Bhuvnesh Rathore, Baseem Khan, Hassan Haes Alhelou, Pierluigi Siano
Summary: This article introduces an algorithm for protecting transmission lines, which detects and locates faults using k-means clustering and weighted k-nearest neighbor (k-NN) regression. The algorithm synchronizes and samples three-phase current signals, computes cumulative differential sum (CDS), and uses various case studies to validate its robustness.
IEEE TRANSACTIONS ON INDUSTRIAL INFORMATICS
(2021)
Article
Computer Science, Artificial Intelligence
Benqiang Wang, Shunxiang Zhang
Summary: The study proposes a new locally adaptive k-nearest centroid neighbour classification method based on average distance, which improves classification performance by finding nearest centroid neighbours to determine k neighbours and deriving discrimination classes with different k values based on the number and distribution of neighbours, resulting in better performance compared to other state-of-the-art KNN algorithms.
CONNECTION SCIENCE
(2022)
Article
Computer Science, Artificial Intelligence
Simone Disabato, Manuel Roveri
Summary: Tiny machine learning (TML) is a new research area focused on designing machine and deep learning techniques for embedded systems and IoT devices. This article introduces a TML for concept drift (TML-CD) solution, which utilizes deep learning feature extractors and a k-nearest neighbors (k-NNs) classifier to adapt to changes in the data-generating process. Experimental results demonstrate the effectiveness of the proposed solution.
IEEE TRANSACTIONS ON NEURAL NETWORKS AND LEARNING SYSTEMS
(2022)
Article
Computer Science, Information Systems
Yibang Ruan, Yanshan Xiao, Zhifeng Hao, Bo Liu
Summary: The paper introduces a nearest-neighbor search model for distance metric learning (NNS-DML), which constructs metric optimization constraints by searching different optimal nearest-neighbor numbers for each training instance. This model reduces the influence of irrelevant features on similar and dissimilar instance pairs and develops a k-free nearest-neighbor model for classification problems. Extensive experiments show that NNS-DML outperforms state-of-the-art distance metric learning methods.
INFORMATION SCIENCES
(2021)
Article
Computer Science, Artificial Intelligence
Shinan Lang, Fangyi Chen, Yiheng Cai
Summary: This paper proposes a highly transparent material classification method based on the imaging model of a time-of-flight (ToF) camera, using features such as refractive index, reflectivity, and transmissivity. The experimental results show that the classification accuracy of this method reaches 94.1% in transparent material classification.
MACHINE VISION AND APPLICATIONS
(2023)
Article
Automation & Control Systems
Atsutake Kosuge, Keisuke Yamamoto, Yukinori Akamine, Takashi Oshima
Summary: This article presents an FPGA-based ICP accelerator which accelerates the object-pose estimation for picking robots through algorithm-level and hardware-level techniques, achieving significant improvement in picking throughput.
IEEE TRANSACTIONS ON INDUSTRIAL ELECTRONICS
(2021)
Article
Engineering, Electrical & Electronic
Xianhao Fan, Jiefeng Liu, Benghui Lai, Yiyi Zhang, Chaohai Zhang
Summary: A model for moisture estimation in transformer oil-paper insulation is proposed using frequency-domain spectroscopy (FDS) and intelligent algorithm, with accuracy and applicability discussed in laboratory and field conditions.
IEEE TRANSACTIONS ON INSTRUMENTATION AND MEASUREMENT
(2021)
Article
Environmental Sciences
Ghasem GhorbanzadehZafarani, Samaneh Karbalaei, Wafaa M. Al-Attar, Reza Golshani, Farhad Hosseini Tayefeh, Arezoo Ashrafizadeh
Summary: Low concentrations of pesticides were found in water samples in the Miankaleh wetland, indicating a low level of aquatic pollution.
MARINE POLLUTION BULLETIN
(2023)
Article
Computer Science, Artificial Intelligence
Jianhua Xia, Jinbing Zhang, Yang Wang, Lixin Han, Hong Yan
Summary: Watershed clustering is a clustering method based on watershed algorithm that can automatically determine the number of clusters in a dataset. In order to handle datasets with multiple dimensions and nonlinear structures, enhancements like KNNG, shared nearest neighbor method, and Pauta Criterion have been introduced. This approach, WC-KNNG-PC, has shown successful performance in clustering various dimensional and complex datasets with heterogeneous density and diverse shapes.
PATTERN RECOGNITION
(2022)
Article
Mathematics
Leilei Liu, Guoyan Zhao, Weizhang Liang
Summary: Slope instability can have catastrophic consequences, but predicting stability is challenging. Machine learning algorithm OPFk-NN shows promising results in slope stability prediction and provides valuable guidance for analysis and risk management.
Article
Geosciences, Multidisciplinary
Mojtaba Zeraatpisheh, Eduardo Leonel Bottega, Esmaeil Bakhshandeh, Hamid Reza Owliaie, Ruhollah Taghizadeh-Mehrjardi, Ruth Kerry, Thomas Scholten, Ming Xu
Summary: The study identified management zones through soil quality assessment and found different soil quality grades within each zone. The research provides a framework to investigate the homogeneity of delineated management zones in terms of soil quality.
Article
Geosciences, Multidisciplinary
Mojtaba Zeraatpisheh, Younes Garosi, Hamid Reza Owliaie, Shamsollah Ayoubi, Ruhollah Taghizadeh-Mehrjardi, Thomas Scholten, Ming Xu
Summary: In this study, the performance of predicting soil organic carbon (SOC) in an arid agroecosystem in Iran using different datasets and machine learning algorithms was compared. The results showed that the Cubist model performed the best with the MCC dataset and the combined dataset of MCC and remote sensing time-series (RST), while the RF model showed better results for the RST dataset. Soil properties were found to be the main factors influencing SOC variation in the MCC and combined datasets, while NDVI was the most controlling factor in the RST dataset. The study suggested that time-series vegetation indices may not significantly improve SOC prediction accuracy, but combining MCC and RST datasets could produce SOC spatial maps with lower uncertainty.
Article
Plant Sciences
Somayeh Emami, Hossein Ali Alikhani, Ahmad Ali Pourbabaee, Hassan Etesami, Fereydoon Sarmadian, Babak Motesharezadeh, Ruhollah Taghizadeh-Mehrjardi
Summary: The study investigated the potential of fluorescent pseudomonads strains in the rhizosphere and endophyte of wheat plants to reduce phosphorus fertilizer application and improve plant traits. The results showed that the combined use of Pseudomonas strains with different levels of phosphorus fertilizer increased wheat growth and yield. Furthermore, the application of these strains also increased soil enzyme activity. This study emphasizes the importance of using chemical-biological fertilizer packages for improving phosphorus nutrition and grain yield in agricultural systems.
JOURNAL OF SOIL SCIENCE AND PLANT NUTRITION
(2022)
Article
Environmental Sciences
Ruhollah Taghizadeh-Mehrjardi, Hossein Khademi, Fatemeh Khayamim, Mojtaba Zeraatpisheh, Brandon Heung, Thomas Scholten
Summary: This study tested and evaluated multiple base learners and model averaging techniques for predicting soil properties in central Iran. The results showed that model averaging approaches can improve the predictive accuracy for soil properties, with different techniques performing better for different soil attributes.
Article
Geosciences, Multidisciplinary
Khadijeh Taghipour, Mehdi Heydari, Yahya Kooch, Hassan Fathizad, Brandon Heung, Ruhollah Taghizadeh-Mehrjardi
Summary: Soil quality, one of the most important characteristics of soil, is crucial for sustainable soil management and evaluating soil degradation. This study aims to assess the impacts of deforestation on soil quality in Iran using a digital soil mapping approach. The results show that the soil quality in the protected forested area is significantly higher than the degraded/deforested area. Machine learning techniques, particularly the Random Forest model, outperform geostatistical approaches in mapping soil quality. This study provides a framework for assessing the impacts of deforestation on soil patterns, which can inform land use planning and forest resource management strategies.
Article
Environmental Sciences
Fatemeh Cheshmberah, Ali A. Zolfaghari, Ruhollah Taghizadeh-Mehrjardi, Thomas Scholten
Summary: Soil Particle Size Distribution (PSD) is a fundamental property that affects soil hydraulic properties and structure. This study evaluated different mathematical models for predicting PSD and used Random Forest to determine the relationship between covariates and the best models' parameters. The results showed that different models performed better for different particle sizes, and the Jaky model performed well in predicting soil particle fractions. The study also demonstrated the potential of combining PSD models and digital soil mapping techniques for spatial distribution analysis.
GEOCARTO INTERNATIONAL
(2022)
Article
Environmental Sciences
Ataollah Shirzadi, Himan Shahabi, Kamal Nabiollahi, Ruhollah Taghizadeh-Mehrjardi, Ivan Lizaga, John J. Clague, Sushant K. Singh, Fariba Golmohamadi, Anuar Ahmad
Summary: This study used machine learning techniques to predict soil erodibility in the Dehgolan region of Iran and compared the performance of five algorithms. The Gaussian Processes model showed the highest prediction accuracy and is valuable for studying soil erodibility in areas with similar climate and soil characteristics.
GEOCARTO INTERNATIONAL
(2022)
Article
Geosciences, Multidisciplinary
Zahra Sohrabizadeh, Hamid Sodaeizadeh, Mohammad Ali Hakimzadeh, Ruhollah Taghizadeh-Mehrjardi, Mohammad Javad Ghanei Bafghi
Summary: This study evaluates the spatial distribution and concentration of heavy metals in soil samples from the Kushk Mine in Bafgh, Iran. Hierarchical clustering analysis, principal component analysis, and spatial distribution patterns were used to assess the distribution of elements in the area. The analysis reveals that heavy metals can be divided into two groups, with lead, cadmium, zinc, and copper influenced by anthropogenic and lithogenic pollution, and iron and manganese impacted by both factors. Higher concentrations of heavy metals were found in the south of the mine and near the tailings.
GEOSCIENCE DATA JOURNAL
(2023)
Article
Environmental Sciences
Tom Broeg, Michael Blaschek, Steffen Seitz, Ruhollah Taghizadeh-Mehrjardi, Simone Zepp, Thomas Scholten
Summary: This study tests the transferability of soil organic carbon (SOC) models for cropland soils using five different types of covariates. The results show that satellite and combined models are transferable, but their accuracy declines. Additionally, mixed-data models significantly improve the accuracies of satellite, terrain, and combined models, while they have no effect on climate models and decrease the models based on soil covariates.
Article
Agronomy
Sedigheh Maleki, Alireza Karimi, Amin Mousavi, Ruth Kerry, Ruhollah Taghizadeh-Mehrjardi
Summary: This study aims to delineate soil management zones (MZs) based on different soil properties using machine learning methods, in order to achieve sustainable agricultural production by maximizing yields and minimizing environmental damage. A random forest model was applied to map soil properties based on environmental covariates, and the study identified four different MZs according to relationships between soil properties and environmental covariates. The ranking of zones in terms of soil fertility was MZ4 > MZ1 > MZ3 > MZ2 based on the investigated soil properties and the soil quality (SQ) map.
Article
Environmental Studies
Yuxin Ma, Budiman Minasny, Valerie Viaud, Christian Walter, Brendan Malone, Alex McBratney
Summary: Soil organic carbon (SOC) redistribution plays a significant role in affecting soil quality. This study introduces a coupled-model combining RothPC-1 and a soil redistribution model to simulate SOC changes in the Lower Hunter Valley area. Results show that erosion is mainly predicted in upslope areas and deposition in low-lying areas. The study emphasizes the importance of considering soil redistribution in SOC dynamics modeling to avoid overestimation of SOC stocks.
Article
Environmental Studies
Masoud Zolfaghari Nia, Mostafa Moradi, Gholamhosein Moradi, Ruhollah Taghizadeh-Mehrjardi
Summary: Spatial variability of soil properties is critical for soil resource planning, management, and exploitation. Different digital soil mapping models were used to estimate soil physicochemical properties in Maroon riparian forests and agricultural lands. The random forest model provided the best estimation for pH, nitrogen, potassium, and bulk density, while the cubist regression tree was more accurate for organic carbon, C:N ratio, phosphorous, and clay. Artificial neural networks showed the best results for calcium carbonate, sand, and silt contents. Geospatial information such as terrain and climate parameters, as well as satellite images, can be effectively used for soil property mapping. Specific machine learning models should be used for each soil property to ensure highly accurate maps.
Article
Multidisciplinary Sciences
Alexandre M. J-C. Wadoux, Mercedes Roman Dobarco, Brendan Malone, Budiman Minasny, Alex B. McBratney, Ross Searle
Summary: This article introduces a new dataset of high-resolution gridded total soil organic carbon content data across Australia. The dataset includes six maps of soil organic carbon content at two resolutions and provides uncertainty estimates. The maps were obtained through interpolation of organic carbon measurements and validation showed small errors and adequate prediction uncertainty. These soil carbon maps are important for monitoring carbon stock changes and assessing the influence of climate change, land management, and greenhouse gas offset.
Article
Soil Science
Ladan Heydari, Hossein Bayat, Fereydoon Sarmadian
Summary: Photosynthetic gas exchange indices can be used to assess spatial variation of crop growth and yield on the field. This study evaluated the spatial relationships between soil water retention curve parameters, soil properties, and photosynthetic gas exchange indices in a wheat field. The results showed that soil variables explained a large part of the total variation in photosynthetic rate and gas exchange parameters. The study also highlighted the potential use of interpolated maps to inform site-specific management decisions and improve wheat yield.
SOIL & TILLAGE RESEARCH
(2023)
Article
Ecology
Mercedes Roman Dobarco, Alexandre M. J-C. Wadoux, Brendan Malone, Budiman Minasny, Alex B. McBratney, Ross Searle
Summary: This study analyzed the soil organic carbon (SOC) composition in Australia, and found that it consists of three fractions: mineral-associated organic carbon (MAOC), particulate organic carbon (POC), and pyrogenic organic carbon (PyOC). These fractions have distinct turnover rates and are influenced differently by different soil environments.