Automatic Detection of Landslides Based on Machine Learning Framework

The rising machine learning (ML) models have become the preferred way for landslide detection based on remote sensing images, but the performance of these models in a sample-free area are rarely concerned in many studies. In this study, we used a cross-validation method (training model in one area and validation in another) to compare the model portability of trained ML models applied in an “off-site” area, as a consideration of the landslide detection ability of these models in sample-free areas. We integrate nighttime light imagery, multi-seasonal optical Landsat time-series and digital elevation data, and we employed support vector machines (SVM), artificial neural networks (ANN) and random forest (RF) models to classify the satellite imagery and identify landslides. Samples of two scenarios generated from two subareas of the Jiuzhaigou disaster-stricken region are used for the cross-application and accuracy evaluation of three ML models. The results revealed that when the trained models are applied in areas outside those in which they were developed, the landslide identification accuracy of these three models has declined. Especially for the SVM and ANN models, the accuracy is greatly reduced and there appears a seriously imbalanced user’s and producer’s accuracy. However, although the performance of the RF model is lower than that of SVM and ANN models in their local area, the RF model exhibits stable portability, and retains the original performance and achieves a satisfactory balance between overestimation and underestimation in “off-site” areas. An additional validation from a new area proved that the landslide detection performance of the RF model with stable portability is higher than that of the SVM and ANN models in “off-site” areas. The results suggest that evaluating the model portability through cross-application can be a useful way to determine the most suitable model for landslide detection in “off-site” areas with a similar geographic environment to model development areas, so as to maximize the accuracy of landslide detection based on limited samples.

Accurate prediction of significant wave height (SWH) is central to coastal ocean dynamics, wave–climate assessment, and operational marine forecasting, yet many high-performing machine-learning (ML) models remain opaque and weakly connected to underlying wave physics. We propose an explainable, feature engineering-guided ML framework for coastal SWH prediction that combines extremal wave statistics, temporal descriptors, and SHAP-based interpretation. Using 30 min buoy observations from a high-energy, wave-dominated coastal site off Australia’s Gold Coast, we benchmarked seven regression models (Linear Regression, Decision Tree, Random Forest, Gradient Boosting, Support Vector Regression, K-Nearest Neighbors, and Neural Networks) across four feature sets: (i) Base (Hmax, Tz, Tp, SST, peak direction), (ii) Base + Temporal (lags, rolling statistics, cyclical hour/month encodings), (iii) Base + a physics-informed Wave Height Ratio, WHR = Hmax/Hs, and (iv) Full (Base + Temporal + WHR). Model skill is evaluated for full-year, 1-month, and 10-day prediction windows. Performance was assessed using R2, RMSE, MAE, and bias metrics, with the Technique for Order Preference by Similarity to Ideal Solution (TOPSIS) employed for multi-criteria ranking. Inclusion of WHR systematically improves performance, raising test R2 from a baseline range of ~0.85–0.95 to values exceeding 0.97 and reducing RMSE by up to 86%, with a Random Forest|Base + WHR configuration achieving the top TOPSIS score (1.000). SHAP analysis identifies WHR and lagged SWH as dominant predictors, linking model behavior to extremal sea states and short-term memory in the wave field. The proposed framework demonstrates how embedding simple, physically motivated features and explainable AI tools can transform black-box coastal wave predictors into transparent models suitable for geophysical fluid dynamics, coastal hazard assessment, and wave-energy applications.

Hospital-acquired complications (HACs) have an adverse impact on patient recovery by impeding their path to full recovery and increasing healthcare costs. The aim of this study was to create a HAC risk prediction machine learning (ML) framework using hospital administrative data collections within North Metropolitan Health Service (NMHS), Western Australia. A retrospective cohort study was performed among 64,315 patients between July 2020 to June 2022 to develop an automated ML framework by inputting HAC and the healthcare site to obtain site-specific predictive algorithms for patients admitted to the hospital in NMHS. Univariate analysis was used for initial feature screening for 270 variables. Of these, 77 variables had significant relationship with any HAC. After excluding non-contemporaneous data, 37 variables were included in developing the ML framework based on logistic regression (LR), decision tree (DT) and random forest (RF) models to predict occurrence of four specific HACs: delirium, aspiration pneumonia, pneumonia and urinary tract infection. All models exhibited similar performance with area under the curve scores around 0.90 for both training and testing datasets. For sensitivity, DT and RF exceeded LR performance while on average, false positives were lowest for LR-based models. Patient's length of stay, Charlson Index, operation length and intensive care unit stay were common predictors. Integrating ML-based risk detection systems into clinical workflows can potentially enhance patient safety and optimise resource allocation. LR-based models exhibited best performance. We have successfully developed a "real-time" risk prediction model, where patient risk scores are calculated and reviewed daily.

Sidoarjo Regency is one of the areas in East Java Province which has the potential for natural and non-natural disasters. So this disaster needs to be overcome with the role of the Sidoarjo Regency BPBD in efforts to deal with the impact of the disaster. So the Sidoarjo Regency BPBD created the Sidoarjo Response Application 'Sigap' which functions for the community if a disaster occurs or is affected by a disaster in the community. This research uses a qualitative method with a descriptive approach using primary and secondary data obtained through observation, interviews and documentation. The focus of this author's research is the effectiveness of the Sidoarjo Response 'Sigap' application in disaster management in Sidoarjo Regency. The results of this research explain that the Sidoarjo Regency BPBD has carried out its role in efforts to increase the effectiveness of the Sigap Application according to its duties and functions, namely: 1). System Quality 2). Information quality 3). Service Quality. There are obstacles or obstacles to research, namely the activities of village communities, but these obstacles can be minimized through the decision of the District BPBD. Sidoarjo. Supporting Factors are the involvement of all elements of Sidoarjo Regency society in participating in activities provided by agencies related to the Effectiveness of the Sigap Application in Disaster Management in Sidoarjo Regency.

UAV usage and interest have shown great popularity recently, especially for commercial multirotor. It is cheap, easy to use, flexible, and has many capabilities. That is why multirotor is widely chosen in disaster management. As a technology, it possesses certain aspects. Review of how the implementation of multirotor UAV in disaster management is important to gain improvement in the technology and evaluate at how we use it. This paper provides insight into commercial multirotor UAV campaign in disaster management through literature study and personal experience in a related theme. Similar ideas on the development of UAV and its application in disaster management then combine with personal experience was used to identify the advantage, disadvantages, challenges, and opportunities of multirotor drone applications. The result suggested that multirotor UAV has limited capability in conducting disaster management campaigns. The weakness in deploying multirotor is related to the lack of technology which may be overcome in the future as the development in hardware and software technology.

This document is the summary report of the IDB-sponsored system of disaster risk and risk management indicators presented at the World Conference on Disaster Reduction in Kobe, Japan, 2005. The indices estimate disaster risk loss, distribution, vulnerability and management for 12 countries in Latin America and the Caribbean. The objective of this program is to facilitate access to relevant information on disaster risk and risk management by national decision-makers, thus making possible the identification and proposal of effective policies and actions. The system of indicators herein proposed permits the benchmarking of each country in different periods, from 1980 to 2000, and cross-national comparisons in a systematic and quantitative fashion allowing the identification of its essential economic and social characteristics. It assists the move toward a more analytically rigorous and data driven approach to risk management decision-making.

Landslide detection from high-resolution satellite imagery plays a significant role in disaster management. Recently, deep learning has emerged as one of the most powerful tools for landslide detection. However, the existing deep learning models for landslide detection still have room for improvement. In this paper, we propose a novel deep learning model named DCA-Net for the automatic detection of landslides. This model adopts an asymmetric encoder-decoder structure. The core module designed in the DCA-Net is the dual-channel depthwise block, integrating two parallel channels of asymmetric depthwise separable convolutions and residual connections to enlarge the receptive field and improve the performance. Experiments are conducted on the open-source Bijie landslide dataset. Several state-of-the-art networks are also employed for quantitative and qualitative comparisons. The results indicate that our proposed DCA-Net is superior to other deep learning models, which can be well utilized in landslide detection from aerial images.

The radar warning and information system for applications in disaster management (RAWIS) is being designed and constructed in a consortium comprising eight partners. It is intended for civil protection applications and will support emergency teams engaged in complex mission contexts. This paper gives an overview of the overall project and addresses some aspects of the sub-projects.

Raman spectroscopy for esophageal tumor diagnosis and delineation using machine learning and the portable Raman spectrometer

Natural disasters such as cyclones and floods recur frequently in certain parts of the world. In this work, we provide a framework to build an easily deployable disaster management application, over five stages. First, we interview four categories of people to understand current problems and approaches in disaster management. Next, we analyze responses and establish that identified disaster management efforts are hitherto unable to effectively harness existing technology. We accordingly build a guided recommendation toolbox of existing Machine Learning (ML), Internet of Things (IOT), and NLP technologies, that satisfies critical system requirements for the identified efforts; this is qualitatively evaluated by senior data scientists and disaster management researchers, and found to reduce development time, as well as increase reliability and user-friendliness. Finally we provide a model to decentralize disaster management. Our work promotes the development of NLP systems tailored for disaster management, bridging the gap between research and real world applications.

With the increasing need for faster and more accurate diagnosis in cardiology, radiomics presents an innovative approach for assessing medical images and diagnosing clinical conditions. This review aims to highlight the applications of radiomics in the diagnosis of cardiovascular conditions. The development of a radiomic model typically progresses as follows: image acquisition and preprocessing, image segmentation, image processing, feature extraction, feature selection, and machine learning modeling and validation. Image data is commonly obtained from cardiac computed tomography angiography, cardiac magnetic resonance imaging, echocardiography, and nuclear imaging. Using machine learning frameworks such as decision trees, random forests, support vector machines, XGBoost, and deep learning, radiomics-based models demonstrated better performance for diagnosis and prediction of cardiovascular events than models designed using conventional clinical risk factors. Radiomics is applied in plaque and adipose tissue characterization to determine the degree of stenosis or predict plaque rupture. In cardiomyopathies, radiomics is employed to distinguish between healthy and diseased tissues. A notable challenge hindering the integration of radiomics in clinical practice is the lack of standardization of study protocols, including image acquisition and processing. Multiple studies also highlighted the need for high-quality images as well as validation of the radiomics model using data from multiple data collection centers. Findings from this study revealed that, while notable advancements have been recorded in radiology-based diagnosis in cardiology, there is a need for further research effort to harmonize evidence and enable the real-world clinical application of radiomics.

Abstract. Artificial Intelligence (AI) has the potential to play a significant role in disaster risk management – from predicting disasters to optimizing response efforts. However, its application in disaster management also presents significant challenges. These challenges include the need for high-quality and diverse data, compatibility with existing systems and technologies, ethical and social implications, and ongoing research and development. In addition, data privacy and security is a critical issue as the use of AI in disaster management often involves the collection and analysis of sensitive information. Addressing these challenges is crucial to ensuring that AI is developed and used in ways that are fair, equitable, and effective in reducing the impact of disasters. The aim of the paper is to provide an analysis of challenges of artificial intelligence application for disaster risk management.

Advanced data analytics in disaster management: A bibliometric analysis of global research trends, methods, and collaboration networks (2006–2025)

Landslide detection and segmentation are critical for disaster risk assessment and management. However, achieving accurate segmentation remains challenging due to the complex nature of landslide terrains and the limited availability of high-quality labeled datasets. This paper proposes an enhanced U-Net++ model for semantic segmentation of landslides in the Wenchuan region using the CAS Landslide Dataset. The proposed model integrates multi-scale feature extraction and attention mechanisms to enhance segmentation accuracy and robustness. The experimental results demonstrate that ASK-UNet++ outperforms traditional methods, achieving a mean intersection over union (mIoU) of 97.53%, a Dice coefficient of 98.27%, and an overall accuracy of 96.04%. These findings highlight the potential of the proposed approach for improving landslide monitoring and disaster response strategies.

Artificial intelligence (AI)-based emergency response systems have emerged as critical enablers of smart infrastructure safety, offering enhanced real-time decision-making, risk assessment, and disaster mitigation strategies across various domains. This systematic literature review, encompassing 424 eligible studies, investigates the integration of machine learning (ML), deep learning (DL), computer vision, IoT-enabled predictive analytics, and AI-powered robotics in optimizing emergency response mechanisms. The study comprehensively examines AI applications in disaster management, real-time incident detection, healthcare emergency response, industrial hazard prevention, cybersecurity frameworks, and intelligent traffic control, providing a detailed assessment of technological advancements and challenges in AI adoption. The findings reveal that AI has significantly improved predictive accuracy, automated hazard detection, and emergency resource optimization, leading to faster response times, minimized human error, and enhanced situational awareness in crisis management. AI-driven predictive analytics models have enabled early warning systems for earthquakes, floods, and wildfires, facilitating proactive disaster preparedness and risk mitigation. In real-time emergency response, AI-powered computer vision and sensor-based surveillance technologies have improved incident detection, reducing intervention delays and ensuring more efficient allocation of emergency resources. In the healthcare sector, AI-enhanced diagnostic tools, triage automation, and geospatial analytics for ambulance dispatch have streamlined medical crisis management, improving survival rates and reducing treatment delays. Additionally, AI-integrated industrial safety frameworks, robotic automation, and cybersecurity intelligence systems have strengthened workplace hazard prevention, cyber threat detection, and emergency communication resilience, ensuring safer and more secure operational environments. Despite these advancements, several challenges related to interoperability, regulatory constraints, cybersecurity vulnerabilities, algorithmic biases, and ethical concerns persist, hindering large-scale AI adoption in emergency response systems. This review provides a comprehensive synthesis of AI’s transformative role in modern emergency management, offering insights into technological developments, limitations, and policy considerations necessary to enhance AI-driven crisis response strategies and ensure more effective, scalable, and resilient emergency safety infrastructures worldwide.