Nearest-neighbor Algorithm Research Articles

Pneumococcal Vaccination Lowers the Risk of Alzheimer's Disease: A Study Utilizing Data from the IBM® MarketScan® Database.

Previous studies demonstrated an association between influenza vaccination and the likelihood of developing Alzheimer's disease. This study was aimed at assessing whether pneumococcal vaccinations are associated with a lower risk of Alzheimer's disease based on analysis of data from the IBM® MarketScan® Database. Vaccinated and unvaccinated matched cohorts were generated using propensity-score matching with the greedy nearest-neighbor matching algorithm. The conditional logistic regression method was used to estimate the relationship between pneumococcal vaccination and the onset of Alzheimer's disease. There were 142,874 subjects who received the pneumococcal vaccine and 14,392 subjects who did not. The conditional logistic regression indicated that the people who received the pneumococcal vaccine had a significantly lower risk of developing Alzheimer's disease as compared to the people who did not receive any pneumococcal vaccine (OR=0.37; 95%CI: 0.33-0.42; P-value < .0001). Our findings demonstrated that the pneumococcal vaccine was associated with a 63% reduction in the risk of Alzheimer's disease among US adults aged 65 and older.

Penerapan Algoritma Sweep dan Particle Swarm Optimization (PSO) sebagai Alternatif Menentukan Rute Distribusi

&lt;em&gt;&lt;span lang="EN-ID"&gt;One aspect of marketing activities is distribution. In the process of distributing goods, it is important to determine the optimal route that minimize mileage and reduce costs. This study aims to provide alternative solutions in determining distribution routes with the shortest distance which has implications for shorter travel times and lower costs. This research adapts the Capacitated Vehicle Routing Problem (CVRP) model with the approach of sweep and Particle Swarm Optimization (PSO) algorithm to determine the route. To generate a comparison route, we use the Nearest Neighbor (NN) algorithm. The result was that 100 agents were divided into 6 clusters and the total distance of the PSO-generated route is 218.115 units or 85.70% of the route distance generated by Nearest Neighbor algorithm&lt;/span&gt;&lt;/em&gt;&lt;span lang="EN-ID"&gt;.&lt;/span&gt;

The Role and Strategy of Big Data in Promoting Local Tourism Development

Abstract As science and technology advance, the tourism industry is increasingly pivoting towards digital and intelligent transformation. The pivotal role of big data in shaping tourism trends is becoming more apparent. This study leverages big data to explore its impact on tourism development and proposes a big data-based model for this purpose. An influence mechanism model tailored to local tourism’s unique characteristics was developed to guide this analysis. Employing data mining techniques, this research utilizes the association rule model to forecast local tourism trends. Additionally, it applies heat evaluation and hotspot analysis models to examine local tourism activities. Spatial characteristics of local tourism are investigated using the nearest neighbor algorithm. An empirical study conducted in Province H illustrates these concepts in action. Here, the entertainment and leisure segment aligns most closely with the overall clustering characteristics, achieving a Z-score of -200.01. The peak hotness months in Province H are May and October, with hotness indices of 0.0438 and 0.0468, respectively. Entertainment and leisure activities make the most significant contribution to this metric, with an average value of 0.0235. Moreover, the prediction error does not exceed 5%, underscoring the validity and practical utility of this study’s methodological approach in analyzing local tourism development. This provides a robust scientific basis for harnessing data to foster local tourism growth.

Implementing AI Solutions for Advanced Cyber‐Attack Detection in Smart Grid

As the backbone of modern power systems, the smart grid is one of the most critical applications of the Internet of Things. The intelligent electricity system faces various types of unauthorized malicious access, that is, cyber‐attacks. With the development and application of information and communication technologies in traditional power systems, the improvement of physical‐cyber systems in the smart grid also increases. Nowadays, the deployment of defensive measures has surged in response to the growing number of attacks aimed at the smart grid. Therefore, in this paper, we investigate cyber‐attack identification using artificial intelligence‐based models and identify them by estimating the state vector of the electricity network. We accomplish simulations by extracting data from the five‐bus IEEE network to test the effectiveness of the proposed algorithm. False data attack vectors are then injected into the healthy measurements. In this way, to check the detection power of the proposed algorithms, 2,000 measurement samples have been taken from the five‐bus network, half of which are considered healthy data and the other half as manipulated data. After labeling healthy and false data, machine‐learning algorithms such as decision trees and k‐nearest neighbor (KNN) have been used to investigate and identify this type of attack. Comparative analysis of the two proposed algorithms against commonly used methods demonstrates significantly improved accuracy. Specifically, according to the best depth for the decision‐tree algorithm and k for the KNN algorithm, it is drawn with k = 3 and the decision‐tree algorithm with a depth of 9. According to the algorithms proposed in this article, the decision‐tree algorithm with a depth of 9 in p‐value of 0.45 and 0.64 and the nearest neighbor algorithm with k = 3 in p is equal to 0.72, 0.98, and 1 represent better accuracy. Also, the results indicate that the two proposed algorithms have performed much more favorably than other classification methods. Additionally, the detection accuracy increases with higher p‐values for these two algorithms. This problem shows that the detectors can detect false data injection attacks that cause more severe disturbances in the system.

Fatigue Life and Crack Growth Rate Prediction of Additively Manufactured 17-4 PH Stainless Steel using Machine Learning

The present work is focused on machine learning-assisted predictions of the low cycle fatigue behaviour and fatigue crack growth rate (FCGR) of 17- 4 PH SS processed through L-PBF and post-processing. Various machine learning techniques reported in the literature provided a flexible approach for explaining the complex mathematical interrelationship among processing-structure-property of the materials. In the present work, four machine learning (ML) algorithms, such as K- Nearest Neighbor (KNN), Decision Trees (DT), Random Forests (RF), and Extreme Gradient Boosting (XGB) algorithms, are implemented to analyze the Fatigue Crack growth rate (FCGR) of 17-4 PH SS alloy. After optimizing the hyper parameters for these algorithms, the trained models were found to estimate the unseen data as equally well as the trained data. The four tested ML models are compared among each other over the training as well as the testing phase based on their mean squared error and R2 scores. Extreme Gradient Boosting model has performed better for the FCGR predictions providing the least mean squared errors and higher R2 scores compared to other models.

Enhancing fraud detection in auto insurance and credit card transactions: a novel approach integrating CNNs and machine learning algorithms.

Fraudulent activities especially in auto insurance and credit card transactions impose significant financial losses on businesses and individuals. To overcome this issue, we propose a novel approach for fraud detection, combining convolutional neural networks (CNNs) with support vector machine (SVM), k nearest neighbor (KNN), naive Bayes (NB), and decision tree (DT) algorithms. The core of this methodology lies in utilizing the deep features extracted from the CNNs as inputs to various machine learning models, thus significantly contributing to the enhancement of fraud detection accuracy and efficiency. Our results demonstrate superior performance compared to previous studies, highlighting our model's potential for widespread adoption in combating fraudulent activities.

Identification of wood specimens utilizing fs-LIBS and machine learning techniques

We report on the ability to identify wood specimens by utilizing 30 fs Laser Induced Breakdown Spectroscopy (LIBS) in conjunction with machine learning techniques. Ten different wood specimens have been studied. The spectral features were assigned to atomic/ionic and diatomic molecular transitions. The origin of the latter has been explored by investigating the dynamics of the created plume in ambient and argon atmosphere. Principal Component Analysis (PCA) was employed for dimensionality reduction based on the primary LIBS analysis. The principal components formation is grounded on the CN, Ca II, Ca I, and Na, LIBS data. Furthermore, applying the weighted k nearest neighbor (kNN) algorithm led to an accurate identification of the investigated specimens, since the evaluation metrics of specificity value were found to be in the range of 0.96–1.00, while that of accuracy was within 0.93–1.00.

Scalable stellar evolution forecasting

Many astrophysical applications require efficient yet reliable forecasts of stellar evolution tracks. One example is population synthesis, which generates forward predictions of models for comparison with observations. The majority of state-of-the-art rapid population synthesis methods are based on analytic fitting formulae to stellar evolution tracks that are computationally cheap to sample statistically over a continuous parameter range. The computational costs of running detailed stellar evolution codes, such as MESA, over wide and densely sampled parameter grids are prohibitive, while stellar-age based interpolation in-between sparsely sampled grid points leads to intolerably large systematic prediction errors. In this work, we provide two solutions for automated interpolation methods that offer satisfactory trade-off points between cost-efficiency and accuracy. We construct a timescale-adapted evolutionary coordinate and use it in a two-step interpolation scheme that traces the evolution of stars from zero age main sequence all the way to the end of core helium burning while covering a mass range from 0.65 to 300 M⊙. The feedforward neural network regression model (first solution) that we train to predict stellar surface variables can make millions of predictions, sufficiently accurate over the entire parameter space, within tens of seconds on a 4-core CPU. The hierarchical nearest-neighbor interpolation algorithm (second solution) that we hard-code to the same end achieves even higher predictive accuracy, the same algorithm remains applicable to all stellar variables evolved over time, but it is two orders of magnitude slower. Our methodological framework is demonstrated to work on the MESA ISOCHRONES AND STELLAR TRACKS (Choi et al. 2016) data set, but is independent of the input stellar catalog. Finally, we discuss the prospective applications of these methods and provide guidelines for generalizing them to higher dimensional parameter spaces.

Comparative Interrupted Time Series Analysis of Direct Medical Expense and Length of Stay in Elderly Patients with Femoral Neck Fractures Who Underwent Total Hip Arthroplasty and Hemiarthroplasty: A Real World Nationwide Database Study.

The objective of our study was to analyze the postoperative direct medical expenses and hospital lengths of stay (LOS) of elderly patients who had undergone either hemiarthroplasty (HA) or total hip arthroplasty (THA) for femoral neck fractures and to determine the indication of THA by comparing those variables between the 2 groups by time. In this comparative large-sample cohort study, we analyzed data from the 2011 to 2018 Korean National Health Insurance Review and Assessment Service database. The included patients were defined as elderly individuals aged 60 years or older who underwent HA or THA for a femoral neck fracture. A 1:1 risk-set matching was performed on the propensity score, using a nearest-neighbor matching algorithm with a maximum caliper of 0.01 of the hazard components. In comparative interrupted time series analysis, time series were constructed using the time unit of one-quarter before and after 3 years from time zero. For the segmented regression analysis, we utilized a generalized linear model with a gamma distribution and logarithmic link function. A total of 4,246 patients who received THA were matched and included with 4,246 control patients who underwent HA. Although there was no statistically significant difference in direct medical expense and hospital LOS for the first 6 months after surgery, direct medical expenses and hospital LOS in THA were relatively reduced compared to the HA up to 24 months after surgery (p < 0.05). In the subgroup analysis, the THA group's hospital LOS decreased significantly compared to that of the HA group during the 7 to 36 months postoperative period in the 65 ≤ age < 80 age group (p < 0.05). Direct medical expenses of the THA group significantly decreased compared to those of the HA group during the period from 7 to 24 months after surgery in the men group (p < 0.05). When performing THA in elderly patients with femoral neck fractures, the possibility of survival for at least 2 years should be considered from the perspective of medical expense and medical utilization. Additionally, in healthy and active male femoral neck fracture patients under the age of 80 years, THA may be more recommended than HA.

Very severe oligozoospermia with AZFc microdeletion patients may affect intracytoplasmic sperm injection clinical outcomes: A propensity score matching analysis.

To explore whether spermatozoa from AZFc microdeletion patients affect their outcomes of intracytoplasmic sperm injection (ICSI). Eighty-five patients with AZFc microdeletion were recruited. A control group of one hundred and forty patients with severe oligozoospermia but without AZF microdeletion was selected using propensity score matching analysis with a 1:2 nearest neighbor algorithm ratio. The ICSI outcomes of the two groups were compared. AZFc microdeletion had lower rates of normal fertilization (73% vs. 80%, p = 0.17) and high-quality embryos (44% vs. 58%, p = 0.07) than the control group. There was no significant difference in the clinical pregnancy rate, miscarriage rate, and live birth rate between the two groups. When the sperm concentration was <1 million/mL, the AZFc microdeletion group exhibited lower rates of fertilization (71% vs. 80%, p = 0.03), high-quality embryo (44% vs. 58%, p = 0.02), clinical pregnancy (57% vs. 76%, p = 0.02), and live birth (49% vs. 72%, p = 0.01) than the control group. However, if sperm concentration was ≥1 million/mL, no significant differences were found. If the sperm concentration is <1 million/mL, AZFc microdeletion do have a detrimental effect on most outcomes of ICSI.

TSP METHOD USING NEAREST NEIGHBOR ALGORITHM AT PT. J&amp;T EXPRESS IN BANDUNG

Effectiveness and efficiency are very important for an expedition company in optimizing the delivery of goods by a courier. The Traveling salesman problem (TSP) method using the Nearest Neighbor Algorithm can optimize the delivery of goods to all consumer location points with only one visit in one trip. The purpose of this study is to find the shortest route using the TSP method based on the travel distance data from drop point PT. J&amp;T Express Sarijadi Bandung to all consumer points and back to this drop point. This data is processed using Matlab and Excel Solver software based on the Nearest Neighbor Algorithm. The results of this study show that the TSP method produces the shortest route, which is 1,944 meters. The delivery route generated by this method provides travel distance efficiency of 50.09% from the route without the TSP method, which is 3,960 meters. Thus, it is expected that there will also be optimization of time and transportation costs in this delivery.

Optimization of Map Matching Algorithm in Various Road Conditions

Map matching refers to find the real driving trajectory of the vehicle by its GPS data. To solve this problem, most research focuses on improving the accuracy of GPS positioning. Even so, map matching still has a large deviation in complex road sections, such as interchanges, underground tunnels and other areas with high spatial complexity. At the same time, traditional map matching algorithms, such as hidden Markov model (HMM) and nearest neighbor algorithms, have the disadvantage of long running time. Different from the research on improving the accuracy of positioning, this research focuses on optimizing and creating a new algorithm for map matching. In order to solve the accuracy of map matching under complex road conditions, this paper adds a new attribute of altitude to GPS data for the first time, and greatly improves the accuracy of map matching in complex road conditions by constructing a three-dimensional hidden Markov model (3D HMM). In order to reduce the running time of the algorithm, this paper uses graph convolutional neural network as a bridge, integrates a variety of map matching algorithms and constructs a hybrid algorithm. Through experiments, it is found that the running time and accuracy of this hybrid algorithm are better than other algorithms.

Prediction of viral pneumonia based on machine learning models analyzing pulmonary inflammation index scores

ObjectTo obtain Pulmonary Inflammation Index scores from imaging chest CT and combine it with clinical correlates of viral pneumonia to predict the risk and severity of viral pneumonia using a computer learning model. MethodsAll patients with suspected viral pneumonia on CT examination admitted to The Affiliated Traditional Chinese Medicine Hospital of Southwest Medical University from December 2022 to March 2023 were retrospectively selected. The respiratory viruses were monitored by RT-PCR and categorized into patients with viral pneumonia and those with non-viral pneumonia. The extent of lung inflammation was quantified according to the Pulmonary Inflammation Index score (PII). Information on patient demographics, comorbidities, laboratory tests, pathogenetic testing, and radiological data were collected. Five machine learning models containing Random Forest(RF), Radial Basis Function Neural Network (RBFNN), Support Vector Machine (SVM), K Nearest Neighbour Algorithm (KNN), and Kernel Ridge Regression (KRR) were used to predict the risk of onset and severity of viral pneumonia based on the clinically relevant factors or PII. ResultsAmong the five models, the SVM model performed best in ACC (76.75 %), SN (73.99 %), and F1 (72.42 %) and achieved a better area under the receiver operating characteristic curve (ROC) (0.8409) when predicting the risk of developing viral pneumonia. RF had the best overall classification accuracy in predicting the severity of viral pneumonia, especially in predicting pneumonia with a PII classification of grade I, the RF model achieved an accuracy of 98.89%. ConclusionMachine learning models are valuable in assessing the risk of viral pneumonia. Meanwhile, machine learning models confirm the importance in predicting the severity of viral pneumonia through PII. The establishment of machine learning models for predicting the risk and severity of viral pneumonia promotes the further development of machine learning in the medical field.

Optimizing the KNN Algorithm to Diagnose Obstructive Pulmonary Diseases

Introduction: According to the World Health Organization, lung diseases are the third cause of death in the world. These diseases are chronic, so early diagnosis of these diseases is very important. Pulmonary function tests are important tools in examining and monitoring patients with respiratory injuries. This research aimed to optimize the K -Nearest Neighbor algorithm, which facilitates and accelerates self-assessment and interpretation of spirometry test results with higher accuracy. Method: In this study, a method is proposed that improves the limitations of the basic algorithm by optimizing, valuing features, and weighted voting. Using this method, obstructive pulmonary diseases are detected based on the data set of spirometry tests, and general parameters are classified into three categories, namely, asthma, chronic bronchitis, and emphysema. Results: In determining the appropriate method for calculating the data distance, the Minkowski method was chosen, and by applying the coefficients of the feature values, the accuracy of the classification increased. Weighted voting was done in the final part of the algorithm based on the Gaussian kernel, based on which a constant performance was obtained for changing the parameter of the number of neighbors. The results of the evaluations were carried out in the form of mutual validation. 95.4% accuracy and 93.2% precision were obtained in 3.12 seconds. Conclusion: The use of machine learning algorithms can be effective in the analysis of medical data. Therefore, in this study, these approaches were used to provide a new method of classification, so that the proposed algorithm could improve the basic method, and also, had better accuracy and performance than other previous methods.

Difference in Quality and Volatile Flavor Compounds of Zingiber officinale Roscoe with Different Drying Methods

In order to study the effect of different drying methods on the quality of ginger (Zingiber officinale Roscoe) in Guizhou Province, the study comparatively analyzed the edible quality, functional components, and volatile flavor compounds of dried ginger. The results revealed that vacuum freeze-drying (VFD) was superior to vacuum microwave drying (WAD), hot air drying (HAD), and natural drying (ND) in terms of retaining ginger color change, rehydration rate, flavonoids, and ginger spicy element. The polyphenol retention in VFD ginger was second only to that in WAD ginger. A total of 86 volatile flavor compounds were detected in fresh ginger and all four drying samples. Ester and aldehyde compounds were identified as the main flavor components in dried ginger, with variations observed among the different drying methods. Principal component analysis (PCA) and the nearest neighbor algorithm successfully distinguished the ginger samples treated with different drying methods. The aroma activity value (OAV) was calculated based on the threshold, and 12 volatile flavor compounds with OAV ≥1 were identified as the key aroma components of ginger. The aroma profiles of ginger were generally similar across the four different drying methods, among which WAD ginger exhibited distinct flavor characteristics associated with drying. This study employed physicochemical properties, active components, and GC-IMS to evaluate the effect of different drying methods on ginger, providing valuable insights for the processing and flavor quality control of drying ginger.

Smoking Classification Using Novel Plasma Cytokines by implementing Machine Learning and Statistical Methods.

Smoking is a major cause of premature and preventable death. Tobacco exposure has a detrimental effect on many organs and contributes to multiple diseases including chronic obstructive pulmonary disease (COPD), cardiovascular disease, cancer, and diabetes. Cytokines are inflammatory biomarkers that are mechanistically associated with smoking. Machine Learning algorithms allow for the quantitative assessment of the contributions of individual cytokines to tobacco-related diseases. The mapping of cytokines to disease can facilitate and direct treatment modalities. By the application of k Nearest Neighbor (k-NN) and Random Forest machine learning algorithms on 63 plasma cytokines we have demonstrated the classification of smoking. To ensure optimal results, performance improvement techniques such as k-fold cross validation and hyper parameter tuning are employed. Separability efficiency achieved by the models is evaluated using the Area Under the Receiver Operating Characteristic (AUROC) metric. The most significant cytokines that enabled the classification are identified and presented. The statistically significant difference for AUROC score of k-NN and Random Forest has been ascertained using the 2-sample independent t test. A reasonably good classification performance was achieved by k-NN algorithm with an AUROC metric of .87, and a 95% CI of (.823,.917). Random forest exceeded k-NN algorithm's performance, with a perfect AUROC score of 1 and a 95% CI of (1,1). From among the ten most prominent cytokines that contributed to the classification, the ones common to both algorithms are: LIF, IL22, G-CSF/CSF-3, TRAIL. AUROC scores for k-NN and Random Forest are significantly different (p-value = 5.105e-16). The discovery and transference of biomarkers such as cytokines from the platform of molecular investigation to clinical practice, can facilitate precision medicine-based therapeutic interventions.

Machine Learning Models for Predicting Air Pollution

Abstract: The air quality monitoring system collects data of pollutants from different location to maintain optimum air quality. In the current situation, it is the critical concern,. The introduction of hazardous gases into the atmosphere from industrial sources, vehicle emissions, etc. pollutes the air. Today, the amount of air pollution in large cities has surpassed the governmentset air quality index value and reached dangerous levels. It has a significant effect on a human health. The prediction of air pollution can be done by the Machine Learning (ML) algorithms. Machine Learning (ML) combines statistics and computer science to maximize the prediction power. ML is used in order to calculate the Air Quality Index. Various sensors and an Arduino Uno microcontroller are utilized to collect the dataset. Then by using K- Nearest Neighbor (KNN) algorithm, the air quality is predicted.

Voting Scheme Nearest Neighbors by Difference Distance Metrics Measurement

K-Nearest Neighbor (KNN) is a widely used method for both classification and regression cases. This algorithm, known for its simplicity and effectiveness, relies primarily on the Euclidean formula for distance metrics. Therefore, this study aimed to develop a voting model where observations were made using different distance calculation formulas. The nearest neighbors algorithm was divided based on differences in distance measurements, with each resulting model contributing a vote to determine the final class. Consequently, three methods were proposed, namely k-nearest neighbors (KNN), Local Mean-based KNN, and Distance-Weighted neighbor (DWKNN), with an inclusion of a voting scheme. The robustness of these models was tested using umbilical cord data characterized by imbalance and small dataset size. The results showed that the proposed voting model for nearest neighbors consistently improved performance by an average of 1-2% across accuracy, precision, recall, and F1 score when compared to the conventional non-voting method.

Path Planning of an Electric Vehicle for Logistics Distribution Considering Carbon Emissions and Green Power Trading

As environmental awareness continues to grow and government policies provide incentives, electric vehicles (EVs) are becoming more widely used in logistics distribution. Considering green power trading and carbon emissions, this paper addresses the green vehicle routing problem (GVRP) and constructs an electric vehicle path model with time windows to minimize the total cost. To solve the model, a hybrid adaptive genetic algorithm (HAGA) is proposed. An improved nearest-neighbor algorithm is adopted to improve the quality of the initial population, and the adaptive crossover and mutation operators are introduced to achieve the better solution. In addition, based on the Schneider case, HAGA is used to solve the models with and without considering green power trading separately, and the results show that considering green power trading can reduce the total cost by 3.22% and carbon emissions by 23.38 kg. Finally, the experimental simulations further prove that with the increase in case size, HAGA can effectively reduce total cost. And it is beneficial for the popularization of electric vehicles in logistics distribution.

A Weighted k-Nearest-Neighbors-Based Spatial Framework of Flood Inundation Risk for Coastal Tourism—A Case Study in Zhejiang, China

Flood inundation causes socioeconomic losses for coastal tourism under climate extremes, progressively attracting global attention. Predicting, mapping, and evaluating the flood inundation risk (FIR) is important for coastal tourism. This study developed a spatial tourism-aimed framework by integrating a Weighted k Nearest Neighbors (WkNN) algorithm, geographic information systems, and environmental indexes, such as precipitation and soil. These model inputs were standardized and weighted using inverse distance calculation and integrated into WkNN to infer the regional probability and distribution of the FIR. Zhejiang province, China, was selected as a case study. The evaluation results were mapped to denote the likelihood of an FIR, which was then validated by the historical Maximum Inundation Extent (MIE) extracted from the World Environment Situation Room. The results indicated that 80.59% of the WkNN results reasonably confirmed the MIE. Among the matched areas, 80.14%, 90.13%, 65.50%, and 84.14% of the predicted categories using WkNN perfectly coincided with MIE at high, medium, low, and very low risks, respectively. For the entire study area, approximately 2.85%, 64.83%, 10.8%, and 21.51% are covered by a high, medium, low, and very low risk of flood inundation. Precipitation and elevation negatively contribute to a high-medium risk. Drainage systems positively alleviate the regional stress of the FIR. The results of the evaluation illustrate that in most inland areas, some tourism facilities are located in high-medium areas of the FIR. However, most tourism facilities in coastal cities are at low or very low risk, especially from Hangzhou-centered northern coastal areas to southern Wenzhou areas. The results can help policymakers make appropriate strategies to protect coastal tourism from flood inundation. Moreover, the evaluation accuracy of WkNN is higher than that of kNN in FIR. The WkNN-based framework provides a reasonable method to yield reliable results for assessing FIR. The framework can also be extended to other risk-related research under climate change.