Kernel Density Estimation Technique Research Articles

Combining LULC data and agricultural statistics for A better identification and mapping of High nature value farmland: A case study in the veneto Plain, Italy

High Nature Value farmland (HNVf) are characterised by high naturalness of land cover and low intensity of agricultural practices. They are essential for biodiversity conservation in rural environments, and their presence is an important indicator of the effectiveness of different EU policies that aim to support biodiversity in agricultural areas. Consequently, their identification, protection and implementation is of strategic importance. Previous studies have analysed the landscape along an urban-rural-natural gradient, beginning with Land Use and Land Cover (LULC) maps. Building up from this baseline, this paper integrates spatial analysis methods with statistical data related to agricultural practices and their intensity, with the aim of mapping and assessing HNVf in a portion of the Veneto Plain, north-east Italy. In particular, this paper presents a methodology for the identification of HNVf applied to two datasets: (i) the first encompassing only LULC data and (ii) the second encompassing also statistical data on agricultural practices. The aim is to demonstrate how this additional information improves the identification of HNVf.In the first step, a Kernel Density Estimation (KDE) technique is applied to a reclassified LULC map, in order to calculate continuous intensity indicators. A Principal Components Analysis and an ISODATA Cluster Analysis are then performed respectively to remove redundant information and to identify the different landscape structures of the study area. The second analysis follows the same steps, with the difference that LULC intensity indicators are analysed in combination with data on crop rotations, irrigation and livestock from a census survey.The first analysis returns a map of the landscape driven only by different intensities of land use. The second returns a map where the statistics on agricultural practices allow for a better characterisation of the natural value of the landscape. Agricultural statistics improved the results, since they allow the discrimination of lower intensity clusters within the cultivated areas, which are traditionally excluded from HNVf by considering only the land cover. The comparison between the results of the two analyses shows that the combined use of the agricultural statistics determines a more detailed representation of the study area, that allows a better differentiation of the agricultural areas between HNVf and non-HNVf, leading to an improvement of the HNVf identification methodology. The benefit of using additional information can be therefore of interest for territorial planning, with the ultimate aim of promoting biodiversity conservation.

A fast background model using kernel density estimation and distance transform

Background modelling is a key factor for foreground detection, which is imperative for people counting in a dynamic environment. In this paper, a background model with improved fast kernel density estimation technique while adopting distance transform is proposed for people counting. The kernel density estimation background model is improved by early-break method and LUT. Distance transform adopted is used to separate people who are closed together as one blob. It has been tested in a 2.6 GHz Intel Core computer with 25 fps on 432 × 240 images without code optimisation. The result shows that the improved method can be used in real-time processing and has a good result.

Data Analytics for Performance Evaluation Under Uncertainties Applied to an Industrial Refrigeration Plant

Artificial intelligence has bounced into industrial applications contributing several advantages to the field and have led to the possibility to open new ways to solve many actual problems. In this paper, a data-driven performance evaluation methodology is presented and applied to an industrial refrigeration system. The strategy takes advantage of the Multivariate Kernel Density Estimation technique and Self-Organizing Maps to develop a robust method, which is able to determine a near-optimal performance map, taking into account the system uncertainties and the multiple signals involved in the process. A normality model is used to detect and filter non-representative operating samples to subsequently develop a reliable performance map. The performance map allows comparing the plant assessment under the same operating conditions and permits to identify the potential system improvement capabilities. To ensure that the resulting evaluation is trustworthy, a robustness strategy is developed to identify either possible new operation conditions or abnormal situations in order to avoid uncertain assessments. Furthermore, the proposed approach is tested with real industrial plant data to validate the suitability of the method.

Layer-by-Layer Enhancement Strategy of Favorable Features of the Deep Belief Network for Industrial Process Monitoring

Recently, owing to its powerful feature extraction capability, deep learning has been applied to the field of fault detection, in which fault information is reflected in data fluctuation. Deep learning networks transmit fluctuation to the abstract feature layer, and the running state is determined on the basis of the fluctuation of the feature layer. Adjacent layers are connected by weights, whereas the weight value determines the amplification or reduction of input data fluctuation. Thus, abnormal fluctuation information may vanish during layer-by-layer feature extraction and this condition is detrimental to fault detection. In this study, to avoid such an event, the abnormal fluctuation information extracted by the deep belief network is monitored and enhanced layer by layer. Enhancement degree is determined by the fluctuation degree, and the enhanced features are integrated via support vector data description, moving average filter, and kernel density estimation techniques to visualize the current work...

Evaluation of Risk Factor for Children During Drop-off and Pick-up Time around the Primary School in Thailand

Background:Traffic accidents, hazardous behaviors, and caretakers’ opinions have a strong effect on students’ safety during the drop-off and pick-up period. The present research about student drop-off and pick-up behaviors and spatial analysis will be helpful for better understanding school traffic incidents and related risk factors.Objective:To examine the risk factors associated with school drop-off and pick-up in three dimensions: 1) risk zones, 2) risk behaviors, and 3) risk periods.Method:A mixed method approach was employed using 4 sets of data, consisting of 1) the statistics of accidents occurring around primary schools, 2) the student drop-off and pick-up behaviors, 3) the opinions toward risk zones and risk periods, and 4) the data obtained from the survey and interview. The spatial analysis was conducted using kernel density estimation technique. The statistical data analysis was carried out to analyze and compare the questionnaire responses given by the teachers and parents from the primary schools with the highest accident rate in 5 regions of Thailand.Results:Pedestrian crossing points in front of the school gates and blind spots behind the school buildings were considered the risk zones. In the morning, the parents usually dropped the students off at the pavements in front of the school gate. Some of them walked the students to the classrooms. After school, the parents parked their cars and then walked to wait for the students at the meeting points scattered inside or outside the school. The opinions of the teachers and parents regarding the risk periods were found to be significantly different (p< 0.01), which affected the way the teachers and parents looked out for the students’ safety in the morning and after school period.Conclusion:The areas around pedestrian crossing points and school gates require effective transport planning in order to specifically prevent and monitor hazardous incidents. A scattering of drop-off and pick-up points are risk behaviors. Therefore, each school should clearly set up safe drop-off and pick-up points and pay close attention to the safety of students both in the morning and after school. Creating safety for students should start with improving the school environment in order to reduce anxiety and facilitate behavior change.

A prediction-Based VM consolidation approach in IaaS Cloud Data Centers

Recent years have witnessed a rapid growth in exploiting Cloud environments to host and deliver various types of virtualized resources as on-demand services. In order to optimally use Cloud resources, the arrangement of virtual machines (VMs) in physical machines (PMs) must be performed strategically, because the placement of VMs in accordance with the available resources can reduce energy consumption, improve resource utilization and, consequently, can increase companies benefits. However, VMs could have time varying workloads, which leads to degradation of performance and power consumption. Thus, re-configuring the VMs placement is essential. Virtual machine consolidation aims to optimally use the available resources by allocating several virtual machines on a set of physical ones (PMs). To determine the PMs capacities to reallocate VMs, it is important to predict their states based on resource utilization history within each VM, and the past VMs migration traffic. However, a common limitation between existing VM consolidation approaches is the lack of information about the history of (and the future) VM migration traffic. Through this paper, we aim to propose a virtual machine consolidation approach based on the estimation of requested resources and the future VM migration traffic. We exploit the strength of Kernel Density Estimation technique (KDE) as a powerful mean to forecast the future resource usage of each VM, and AKKA toolkit as an actor-based model that allows exchanging useful information about the host’s states. We adopt a weighted-graph representation to model the history of migration traffic between PMs and to design the actor-based topology of the data center. The obtained results show the effectiveness of our approach in terms of total number of migrations and energy consumption.

A negotiation‐based service selection approach using swarm intelligence and kernel density estimation

SummaryNowadays, the cloud computing environment is becoming a natural choice to deploy and provide Web services that meet user needs. However, many services provide the same functionality and high quality of service (QoS) but different self‐adaptive behaviors. In this case, providers' adaptation policies are useful to select services with high QoS and high quality of adaptation (QoA). Existing approaches do not take into account providers' adaptation policies in order to select services with high reputation and high reaction to changes, which is important for the composition of self‐adaptive Web services. In order to actively participate to compositions, candidate services must negotiate their self‐* capabilities. Moreover, they must evaluate the participation constraints against their capabilities specified in terms of QoS and adaptation policies. This paper exploits a variant of particle swarm optimization and kernel density estimation in the selection of service compositions and the concurrent negotiations of their QoS and QoA capabilities. Selection and negotiation processes are held between intelligent agents, which adopt swarm intelligence techniques for achieving optimal selection and optimal agreement on providers' offers. To resolve unknown autonomic behavior of candidate services, we deal with the lack of such information by predicting the real QoA capabilities of a service through the kernel density estimation technique. Experiments show that our solution is efficient in comparison with several state‐of‐the‐art selection approaches.

Data-driven decision making under uncertainty integrating robust optimization with principal component analysis and kernel smoothing methods

This paper proposes a novel data-driven robust optimization framework that leverages the power of machine learning and big data analytics for decision-making under uncertainty. By applying principal component analysis to uncertainty data, correlations between uncertain parameters are effectively captured, and latent uncertainty sources are identified. These data are then projected onto each principal component to facilitate extracting distributional information of latent uncertainties using kernel density estimation techniques. To explicitly account for asymmetric distributions, we introduce forward and backward deviation vectors into the data-driven uncertainty set, which are further incorporated into novel data-driven static and adaptive robust optimization models. The proposed framework not only significantly ameliorates the conservatism of robust optimization, but also enjoys computational efficiency and wide-ranging applicability. Three applications on optimization under uncertainty, including model predictive control, batch production scheduling, and process network planning, are presented to demonstrate the applicability of the proposed framework.

Modeling error PDF optimization based wavelet neural network modeling of dynamic system and its application in blast furnace ironmaking

Abstract In general, the modeling errors of dynamic system model are a set of random variables. The traditional performance index of modeling such as means square error (MSE) and root means square error (RMSE) cannot fully express the connotation of modeling errors with stochastic characteristics both in the dimension of time domain and space domain. Therefore, the probability density function (PDF) is introduced to completely describe the modeling errors in both time scales and space scales. Based on it, a novel wavelet neural network (WNN) modeling method is proposed by minimizing the two-dimensional (2D) PDF shaping of modeling errors. First, the modeling error PDF by the traditional WNN is estimated using data-driven kernel density estimation (KDE) technique. Then, the quadratic sum of 2D deviation between the modeling error PDF and the target PDF is utilized as performance index to optimize the WNN model parameters by gradient descent method. Since the WNN has strong nonlinear approximation and adaptive capability, and all the parameters are well optimized by the proposed method, the developed WNN model can make the modeling error PDF track the target PDF, eventually. Simulation example and application in a blast furnace ironmaking process show that the proposed method has a higher modeling precision and better generalization ability compared with the conventional WNN modeling based on MSE criteria. Furthermore, the proposed method has more desirable estimation for modeling error PDF that approximates to a Gaussian distribution whose shape is high and narrow.

Density-based particle swarm optimization algorithm for data clustering

Abstract Particle swarm optimization (PSO) algorithm is widely used in cluster analysis. However, it is a stochastic technique that is vulnerable to premature convergence to sub-optimal clustering solutions. PSO-based clustering algorithms also require tuning of the learning coefficient values to find better solutions. The latter drawbacks can be evaded by setting a proper balance between the exploitation and exploration behaviors of particles while searching the feature space. Moreover, particles must take into account the magnitude of movement in each dimension and search for the optimal solution in the most populated regions in the feature space. This study presents a novel approach for data clustering based on particle swarms. In this proposal, the balance between exploitation and exploration processes is considered using a combination of (i) kernel density estimation technique associated with new bandwidth estimation method to address the premature convergence and (ii) estimated multidimensional gravitational learning coefficients. The proposed algorithm is compared with other state-of-the-art algorithms using 11 benchmark datasets from the UCI Machine Learning Repository in terms of classification accuracy, repeatability represented by the standard deviation of the classification accuracy over different runs, and cluster compactness represented by the average Dunn index values over different runs. The results of Friedman Aligned-Ranks test with Holm's test over the average classification accuracy and Dunn index values indicate that the proposed algorithm achieves better accuracy and compactness when compared with other algorithms. The significance of the proposed algorithm is represented in addressing the limitations of the PSO-based clustering algorithms to push forward clustering as an important technique in the field of expert systems and machine learning. Such application, in turn, enhances the classification accuracy and cluster compactness. In this context, the proposed algorithm achieves better results compared with other state-of-the-art algorithms when applied to high-dimensional datasets (e.g., Landsat and Dermatology). This finding confirms the importance of estimating multidimensional learning coefficients that consider particle movements in all the dimensions of the feature space. The proposed algorithm can likewise be applied in repeatability matters for better decision making, as in medical diagnosis, as proved by the low standard deviation obtained using the proposed algorithm in conducted experiments.

Adaptive Fingerprinting in Multi-Sensor Fusion for Accurate Indoor Tracking

Indoor Localization and Tracking have become an attractive research topic because of the wide range of potential applications. These applications are highly demanding in terms of estimation accuracy and rise a challenge due to the complexity of the scenarios modeled. Approaches for these topics are mainly based on either deterministic or probabilistic methods, such as Kalman or Particles Filter. These techniques are improved by fusing information from different sources, such as wireless or optical sensors. In this paper, a novel MUlti-sensor Fusion using adaptive fingerprint (MUFAF) algorithm is presented and compared with several multi-sensor indoor localization and tracking methods. MUFAF is mainly divided in four phases: first, a target position estimation (TPE) process is performed by every sensor; second, a target tracking process stage; third, a multi-sensor fusion combines the sensor information; and finally, an adaptive fingerprint update (AFU) is applied. For TPE, a complete environment characterization in combination with a Kernel density estimation technique is employed to obtain object position. A Modified Kalman Filter is applied to TPE output in order to smooth target routes and avoid outliers effect. Moreover, two fusion methods are described in this paper: track-to-track fusion and Kalman sensor group fusion. Finally, AFU will endow the algorithm with responsiveness to environment changes by using Kriging interpolation to update the scenario fingerprint. MUFAF is implemented and compared in a test bed showing that it provides a significant improvement in estimation accuracy and long-term adaptivity to condition changes.

Efficient and fast multi-modal foreground-background segmentation using RGBD data

This paper addresses the problem of foreground and background segmentation. Multi-modal data specifically RGBD data has gain many tasks in computer vision recently. However, techniques for background subtraction based only on single-modality still report state-of-the-art results in many benchmarks. Succeeding the fusion of depth and color data for this task requires a robust formalization allowing at the same time higher precision and faster processing. To this end, we propose to make use of kernel density estimation technique to adapt multi-modal data. To speed up kernel density estimation, we explore the fast Gauss transform which allows the summation of a mixture of M kernel at N evaluation points in O(M+N) time as opposed to O(MN) time for a direct evaluation. Extensive experiments have been carried out on four publicly available RGBD foreground/background datasets. Results demonstrate that our proposal outperforms state-of-the-art methods for almost all of the sequences acquired in challenging indoor and outdoor contexts with a fast and non-parametric operation.

Improvement of the performance of animal crossing warning signs

Introduction: Animal–vehicle collisions (AVCs) can result in serious injury and death to drivers, animals' death, and significant economic costs. However, the cost effectiveness of the majority of AVC mitigation measures is a significant issue. Method: A mobile-based data collection effort was deployed to measure signs under the Utah Department of Transportation's (UDOT) jurisdiction. The crash data were obtained from the UDOT risk management database. ArcGIS was employed to link these two data sets and extract animal-related crashes and signs. An algorithm was developed to process the data and identify AVCs that occurred within sign recognition distance. Kernel density estimation (KDE) technique was applied to identify potential crash hotspots. Results: Only 2% of AVCs occurred within the recognition distance of animal crossing signs. Almost 58% of animal-related crashes took place on the Interstate and U.S. highways, wherein only 30% of animal crossing signs were installed. State routes with a higher average number of signs experienced a lower number of AVCs per mile. The differences between AVCs that occurred within versus outside of sign recognition distance were not statistically significant regarding crash severity, time of crash, weather condition, driver age, vehicle speed, and type of animal. It is more likely that drivers become accustomed to deer crossing signs than cow signs. Conclusions: Based on the historical crash data and landscape structure, with attention given to the low cost safety improvement methods, a combination of different types of AVC mitigation measures can be developed to reduce the number of animal-related crashes. After an in-depth analysis of AVC data, warning traffic signs, coupled with other low cost mitigation countermeasures can be successfully placed in areas with higher priority or in critical areas. Practical applications: The findings of this study assist transportation agencies in developing more efficient mitigation measures against AVCs.

A Priori Assessment of Prediction Confidence for Data-Driven Turbulence Modeling

Although Reynolds-Averaged Navier-Stokes (RANS) equations are still the dominant tool for engineering design and analysis applications involving turbulent flows, standard RANS models are known to be unreliable in many flows of engineering relevance, including flows with separation, strong pressure gradients or mean flow curvature. With increasing amounts of 3-dimensional experimental data and high fidelity simulation data from Large Eddy Simulation (LES) and Direct Numerical Simulation (DNS), data-driven turbulence modeling has become a promising approach to increase the predictive capability of RANS simulations. Recently, a data-driven turbulence modeling approach via machine learning has been proposed to predict the Reynolds stress anisotropy of a given flow based on high fidelity data from closely related flows. In this work, the closeness of different flows is investigated to assess the prediction confidence a priori. Specifically, the Mahalanobis distance and the kernel density estimation (KDE) technique are used as metrics to quantify the distance between flow data sets in feature space. The flow over periodic hills at Re=10595 is used as the test set and seven flows with different configurations are individually used as training set. The results show that the prediction error of the Reynolds stress anisotropy is positively correlated with Mahalanobis distance and KDE distance, demonstrating that both extrapolation metrics can be used to estimate the prediction confidence a priori. A quantitative comparison using correlation coefficients shows that the Mahalanobis distance is less accurate in estimating the prediction confidence than KDE distance. The extrapolation metrics introduced in this work and the corresponding analysis provide an approach to aid in the choice of the data source and to assess the prediction confidence for data-driven turbulence modeling.

Clustering on very small scales from a large sample of confirmed quasar pairs: does quasar clustering track from Mpc to kpc scales?

We present the most precise estimate to date of the clustering of quasars on very small scales, based on a sample of 47 binary quasars with magnitudes of $g<20.85$ and proper transverse separations of $\sim 25\,h^{-1}$\,kpc. Our sample of binary quasars, which is about 6 times larger than any previous spectroscopically confirmed sample on these scales, is targeted using a Kernel Density Estimation technique (KDE) applied to Sloan Digital Sky Survey (SDSS) imaging over most of the SDSS area. Our sample is "complete" in that all of the KDE target pairs with $17.0 \lesssim R \lesssim 36.2\,h^{-1}$\,kpc in our area of interest have been spectroscopically confirmed from a combination of previous surveys and our own long-slit observational campaign. We catalogue 230 candidate quasar pairs with angular separations of $<8\arcsec$, from which our binary quasars were identified. We determine the projected correlation function of quasars ($\bar W_{\rm p}$) in four bins of proper transverse scale over the range $17.0 \lesssim R \lesssim 36.2\,h^{-1}$\,kpc. The implied small-scale quasar clustering amplitude from the projected correlation function, integrated across our entire redshift range, is $A=24.1\pm3.6$ at $\sim 26.6 ~h^{-1}$\,kpc. Our sample is the first spectroscopically confirmed sample of quasar pairs that is sufficiently large to study how quasar clustering evolves with redshift at $\sim 25 ~h^{-1}$ kpc. We find that empirical descriptions of how quasar clustering evolves with redshift at $\sim 25 ~h^{-1}$ Mpc also adequately describe the evolution of quasar clustering at $\sim 25 ~h^{-1}$ kpc.

Using GIS Mapping to Target Public Health Interventions: Examining Birth Outcomes Across GIS Techniques.

With advances in spatial analysis techniques, there has been a trend in recent public health research to assess the contribution of area-level factors to health disparity for a number of outcomes, including births. Although it is widely accepted that health disparity is best addressed by targeted, evidence-based and data-driven community efforts, and despite national and local focus in the U.S. to reduce infant mortality and improve maternal-child health, there is little work exploring how choice of scale and specific GIS visualization technique may alter the perception of analyses focused on health disparity in birth outcomes. Retrospective cohort study. Spatial analysis of individual-level vital records data for low birthweight and preterm births born to black women from 2007 to 2012 in one mid-sized Midwest city using different geographic information systems (GIS) visualization techniques [geocoded address records were aggregated at two levels of scale and additionally mapped using kernel density estimation (KDE)]. GIS analyses in this study support our hypothesis that choice of geographic scale (neighborhood or census tract) for aggregated birth data can alter programmatic decision-making. Results indicate that the relative merits of aggregated visualization or the use of KDE technique depend on the scale of intervention. The KDE map proved useful in targeting specific areas for interventions in cities with smaller populations and larger census tracts, where they allow for greater specificity in identifying intervention areas. When public health programmers seek to inform intervention placement in highly populated areas, however, aggregated data at the census tract level may be preferred, since it requires lower investments in terms of time and cartographic skill and, unlike neighborhood, census tracts are standardized in that they become smaller as the population density of an area increases.

A framework to embed a spatial statistics toolbox in open-source GIS software: kernel density estimation example

It is widely known that geographic information systems (GIS) should include more spatial data analysis (SDA) techniques. The issues of which techniques should be included and how statistical analysis can be integrated with GIS are still widely debated. This paper focuses on the development of a framework that implements R SDA techniques in the uDig. For this purpose, a simple interface is designed between two open-source software applications, uDig and the R statistical software package. The tight coupling strategy is adopted with the RCaller interpreter. Overall, the integration is successfully implemented and tested by users and developers. Fourteen geospatial and four non-geospatial techniques are integrated into uDig successfully, which demonstrates the success of the proposed framework. Among these techniques, the kernel density estimation (KDE) technique is explained with a sample data set to show every step of the implementation. The user tests also prove the success of the integration.

KDE based outlier detection on distributed data streams in multimedia network

Multimedia networks hold the promise of facilitating large-scale, real-time data processing in complex environments. Their foreseeable applications will help protect and monitor military, environmental, safety-critical, or domestic infrastructures and resources. Cloud infrastructures promise to provide high performance and cost effective solutions to large scale data processing problems. This paper focused on the outlier detection over distributed data stream in real time, proposed kernel density estimation (KDE) based outlier detection algorithm KDEDisStrOut in Storm, firstly formalized the problem of outlier detection using the kernel density estimation technique and update the transported data incrementally between the child node and the coordinator node which reduces the communication cost. Then the paper adopted the exponential decay policy to keep pace with the transient and evolving natures of stream data and changed the weight of different data in the sliding window adaptively made the data analysis more reasonable. Theoretical analysis and experiments on Storm with synthetic and real data show that the KDEDisStrOut algorithm is efficient and effective compared with existing outlier detection algorithms, and more suitable for data streams.

Development and application of traffic accident density estimation models using kernel density estimation

Traffic accident frequency has been decreasing in Japan in recent years. Nevertheless, many accidents still occur on residential roads. Area-wide traffic calming measures including Zone 30, which discourages traffic by setting a speed limit of 30 km/h in residential areas, have been implemented. However, no objective implementation method has been established. Development of a model for traffic accident density estimation explained by GIS data can enable the determination of dangerous areas objectively and easily, indicating where area-wide traffic calming can be implemented preferentially. This study examined the relations between traffic accidents and city characteristics, such as population, road factors, and spatial factors. A model was developed to estimate traffic accident density. Kernel density estimation (KDE) techniques were used to assess the relations efficiently. Besides, 16 models were developed by combining accident locations, accident types, and data types. By using them, the applicability of traffic accident density estimation models was examined. Results obtained using Spearman rank correlation show high coefficients between the predicted number and the actual number. The model can indicate the relative accident risk in cities. Results of this study can be used for objective determination of areas where area-wide traffic calming can be implemented preferentially, even if sufficient traffic accident data are not available.

Nonstationarity in the occurrence rate of floods in the Tarim River basin, China, and related impacts of climate indices

Amplification of floods in the Xinjiang, China, has been observed, but reports on their changing properties and underlying mechanisms are not available. In this study, occurrence rates of floods in the Tarim River basin, the largest inland arid river basin in China, were analyzed using the Kernel density estimation technique and bootstrap resampling method. Also analyzed were the occurrence rates of precipitation extremes using the POT (Peak over Threshold)-based sampling method. Both stationary and non-stationary models were developed using GAMLSS (Generalized Additive Models for Location, Scale and Shape) to model flood frequency with time, climate index, precipitation and temperature as major predictors. Results indicated: (1) two periods with increasing occurrence of floods, i.e., the late 1960s and the late 1990s with considerable fluctuations around 2–3 flood events during time intervals between the late 1960s and the late 1990s; (2) changes in the occurrence rates of floods were subject to nonstationarity. A persistent increase of flood frequency and magnitude was observed during the 1990s and reached a peak value; (3) AMO (Atlantic Multidecadal Oscillation) and AO (Atlantic Oscillation) in winter were the key influencing climate indices impacting the occurrence rates of floods. However, NAO (North Atlantic Oscillation) and SOI (South Oscillation Index) are two principle factors that influence the occurrence rates of regional floods. The AIC (Akaike Information Criterion) values indicated that compared to the influence of climate indices, occurrence rates of floods seemed to be more sensitive to temperature and precipitation changes. Results of this study are important for flood management and development of mitigation measures.