Supervised Classification Method Research Articles

Characterization and evaluation of environmental units as a management and conservation strategy of the Palmira Desert (Ecuador)

Globally, deserts are considered as fragile and unique biomes. They have the capacity to provide various ecosystem services, and possess great tourism potential and scientific value, but are also susceptible to modifications or alterations in their geomorphology. Despite the strategic importance of these ecosystems at a global level, there is limited specialized research directed towards the study of deserts in Ecuador. The purpose of this research is to determine the susceptibility to degradation and carrying capacity of environmental units to support tourist and recreational activities, thereby offering the community a sustainable resource over time. The delimitation of environmental units was carried out based on the identification of land use and land cover through the supervised classification method. Multicriteria analysis was used to obtain the areas susceptible to degradation, where the environmental units were evaluated through expert evaluation. Once the results of the previous methodologies were obtained, areas with environmental and tourist potential were identified using the hosting capacity matrix, which is based on anthropic activities developed in the area. Subsequently, a proposal for conservation strategies and activities at the national and local levels was formulated. Five categories of land use and land cover were identified, resulting in 11 environmental units represented in spatial data. It was determined that approximately 31% of the territory is more susceptible to degradation, while the remaining 69% is less susceptible. Based on the susceptibility of these areas, 7 activities and 2 tourist routes were designated to help reduce anthropogenic pressure. Additionally, the community was provided with 5 proposals for conservation strategies at the national level, 4 at the local level, and 4 activities to be developed based on the findings identified during the investigation. This highlights the significant potential that the Palmira Desert has to become a conservation area.

Automatic recognition of excavator working cycles using supervised learning and motion data obtained from inertial measurement units (IMUs)

This paper proposes an automatic method for excavator working cycle recognition using supervised classification methods and motion information obtained from four inertial measurement units (IMUs) attached to moving parts of an excavator. Monitoring and analyzing tasks that have been performed by heavy-duty mobile machines (HDMMs) are significantly required to assist management teams in productivity and progress monitoring, efficient resource allocation, and scheduling. Nevertheless, traditional methods depend on human observations, which are costly, time-consuming, and error-prone. There is a lack of a method to automatically detect excavator major activities. In this paper, a data-driven method is presented to identify excavator activities, including loading, trenching, grading, and idling, using motion information, such as angular velocities and joint angles, obtained from moving parts, including swing body, boom, arm, and bucket. Firstly, a dataset lasting 3 h is collected using a medium-rated excavator. One experienced and one inexperienced operator performed tasks under different working conditions, such as different types of material, swing angle, digging depth, and weather conditions. Four classification methods, including support vector machine (SVM), k-nearest neighbor (KNN), decision tree (DT), and naive Bayes, are off-line trained. The results show that the proposed method can effectively identify excavator working cycles with a high accuracy of 99%. Finally, the impacts of parameters, such as time window, overlapping configuration, and feature selection methods, on the classification accuracy are comprehensively analyzed.

Remote Sensing for Mapping Natura 2000 Habitats in the Brière Marshes: Setting Up a Long-Term Monitoring Strategy to Understand Changes

On a global scale, wetlands are suffering from a steady decline in surface area and environmental quality. Protecting them is essential and requires a careful spatialisation of their natural habitats. Traditionally, in our study area, species discrimination for floristic mapping has been achieved through on-site field inventories, but this approach is very time-consuming in these difficult-to-access environments. Usually, the resulting maps are also not spatially exhaustive and are not frequently updated. In this paper, we propose to establish a complete map of the study area using remote sensors and set up a long-term and regular observatory of environmental changes to monitor the evolution of a major French wetland. This methodology combines three dataset acquisition technologies, airborne hyperspectral and WorldView-3 multispectral images, supplemented by LiDAR images, which we compared to evaluate the difference in performances. To do so, we applied the Random Forest supervised classification methods using ground reference areas and compared the out-of-bag score (OOB score) as well as the matrix of confusion resulting from each dataset. Thirteen habitats were discriminated at level 4 of the European Nature Information System (EUNIS) typology, at a spatial resolution of around 1.2 m. We first show that a multispectral image with 19 variables produces results which are almost as good as those produced by a hyperspectral image with 58 variables. The experiment with different features also demonstrates that the use of four bands derived from LiDAR datasets can improve the quality of the classification. Invasive alien species Ludwigia grandiflora and Crassula helmsii were also detected without error which is very interesting when applied to these endangered environments. Therefore, since WV-3 images provide very good results and are easier to acquire than airborne hyperspectral data, we propose to use them going forward for the regular observation of the Brière marshes habitat we initiated.

Spatio-temporal analysis land use land cover changes in South Kashmir region of North-western Himalayas using Landsat data

This paper presents a comprehensive analysis of the spatial patterns and temporal dynamics of land use and land cover changes in South Kashmir from 2000 to 2022 leveraging remote sensing technologies revealing significant transformations in various land cover classes. The study used maximum likelihood classification, a supervised classification method, to analyze Landsat satellite imagery and identify ten major land use categories. The findings demonstrate notable increases in, barren land by 18.78 km2 (0.35%), built-up areas by 72.28 km2 (1.33%), forests by 274.76 km2 (5.05%), grasslands by 68.06 km2 (1.25%), scrubland by 307.82 km2 (5.66%). horticulture experienced a significant rise of 419.17 km2 (7.70%), Conversely, several land use classes reported decline, agriculture by 757.21 km2 (13.91%), exposed rockmass by 258.58 km2 (4.75%), glaciers and snow by 136.83 km2 (2.51%), and water bodies contracted by 8.23 km2 (0.15%). The primary drivers of land use change in the region are identified as climate change, population growth, and economic factors. Climate change has altered precipitation patterns affecting agricultural productivity and leading to the retreat of glaciers. Population growth and economic reasons, including the rise of horticulture and changes in land use policies, have also played a significant role in shaping the landscape dynamics of South Kashmir. These changes underscore the dynamic nature of land use in South Kashmir, with significant implications for regional planning and environmental management. The study underscores the cost-effectiveness and efficacy of geospatial technologies in conducting spatiotemporal analyses and formulating evidence-based policies for the sustainable management of natural resources.

ChinaSoyArea10m: a dataset of soybean-planting areas with a spatial resolution of 10 m across China from 2017 to 2021

Abstract. Soybean, an essential food crop, has witnessed a steady rise in demand in recent years. There is a lack of high-resolution annual maps depicting soybean-planting areas in China, despite China being the world's largest consumer and fourth-largest producer of soybean. To address this gap, we developed the novel Regional Adaptation Spectra-Phenology Integration method (RASP) based on Sentinel-2 remote sensing images from the Google Earth Engine (GEE) platform. We utilized various auxiliary data (e.g., cropland layer, detailed phenology observations) to select the specific spectra and indices that differentiate soybeans most effectively from other crops across various regions. These features were then input for an unsupervised classifier (K-means), and the most likely type was determined by a cluster assignment method based on dynamic time warping (DTW). For the first time, we generated a dataset of soybean-planting areas across China, with a high spatial resolution of 10 m, spanning from 2017 to 2021 (ChinaSoyArea10m). The R2 values between the mapping results and the census data at both the county and prefecture levels were consistently around 0.85 in 2017–2020. Moreover, the overall accuracy of the mapping results at the field level in 2017, 2018, and 2019 was 77.08 %, 85.16 %, and 86.77 %, respectively. Consistency with census data was improved at the county level (R2 increased from 0.53 to 0.84) compared to the existing 10 m crop-type maps in Northeast China (Crop Data Layer, CDL) based on field samples and supervised classification methods. ChinaSoyArea10m is very spatially consistent with the two existing datasets (CDL and GLAD (Global Land Analysis and Discovery) maize–soybean map). ChinaSoyArea10m provides important information for sustainable soybean production and management as well as agricultural system modeling and optimization. ChinaSoyArea10m can be downloaded from an open-data repository (DOI: https://doi.org/10.5281/zenodo.10071427, Mei et al., 2023).

ANALISIS SPASIAL PERUBAHAN TUTUPAN LAHAN PASCA KEBAKARAN HUTAN DAN LAHAN DI KABUPATEN MUARO JAMBI

Changes in land use and cover play an important role in sustainable natural resource management. Muaro Jambi Regency, which is rich in biodiversity, has ecosystem services for residents in the area, especially plantations which are one of the leading commodities. Forests as the largest land cover in Muaro Jambi Regency have experienced quite rapid changes, especially after forest and land fires occur almost every year. Real-time monitoring and assessment on a spatio-temporal basis is important to determine changes in land cover. This research examines changes in land cover in Sarolangun Regency from Landsat-8 time series images from 2018 to 2023. Land cover is classified into 6 classes using the supervised classification method. The data findings show that forests have experienced a significant decline, while settlements have experienced a very substantial increase. Comprehensive research regarding the vulnerability of periodic land cover changes at local and regional levels must be carried out to maintain sustainable environmental ecosystems.

Standardization with zlog values improves exploratory data analysis and machine learning for laboratory data

Abstract Objectives In the context of exploratory data analysis and machine learning, standardization of laboratory results is an important pre-processing step. Variable proportions of pathological results in routine datasets lead to changes of the mean (µ) and standard deviation (σ), and thus cause problems in the classical z-score transformation. Therefore, this study investigates whether the zlog transformation compensates these disadvantages and makes the results more meaningful from a medical perspective. Methods The results presented here were obtained with the statistical software environment R, and the underlying data set was obtained from the UC Irvine Machine Learning Repository. We compare the differences of the zlog and z-score transformation for five different dimension reduction methods, hierarchical clustering and four supervised classification methods. Results With the zlog transformation, we obtain better results in this study than with the z-score transformation for dimension reduction, clustering and classification methods. By compensating the disadvantages of the z-score transformation, the zlog transformation allows more meaningful medical conclusions. Conclusions We recommend using the zlog transformation of laboratory results for pre-processing when exploratory data analysis and machine learning techniques are applied.

New Strategies in Archaeometric Provenance Analyses of Volcanic Rock Grinding Stones: Examples from Iulia Libica (Spain) and Sidi Zahruni (Tunisia)

Archaeometry can help archaeologists in many ways, and one of the most common archaeometric objectives is provenance analysis. Volcanic rocks are often found in archaeological sites as materials used to make grinding tools such as millstones and mortars or as building materials. Petrographic characterization is commonly applied to identify their main mineralogical components. However, the provenance study of volcanic stones is usually undertaken by comparing geochemical data from reference outcrops using common descriptive statistical tools such as biplots of chemical elements, and occasionally, unsupervised multivariate data analysis like principal component analysis (PCA) is also used. Recently, the use of supervised classification methods has shown a superior performance in assigning provenance to archaeological samples. However, these methods require the use of reference databases for all the possible provenance classes in order to train the classification models. The existence of comprehensive collections of published geochemical analyses of igneous rocks enables the use of the supervised approach for the provenance determination of volcanic stones. In this paper, the provenance of volcanic grinding tools from two archaeological sites (Iulia Libica, Spain, and Sidi Zahruni, Tunisia) is attempted using data from the GEOROC database through unsupervised and supervised approaches. The materials from Sidi Zahruni have been identified as basalts from Pantelleria (Italy), and the agreement between the different supervised classification models tested is particularly conclusive. In contrast, the provenance of the materials from Iulia Libica remained undetermined. The results illustrate the advantages and limitations of all the examined methods.

Analyses of LULC dynamics in a socio-ecological system of the Bale Mountains Eco Region of Southeast Ethiopia.

Analysis of land use and land cover (LULC) change and its drivers and impacts in the biodiversity hotspot of Bale Mountain's socio-ecological system is crucial for formulating plausible policies and strategies that can enhance sustainable development. The study aimed to analyze spatio-temporal LULC changes and their trends, extents, drives, and impacts over the last 48 years in the Bale Mountain social-ecological system. Landsat imagery data from the years 1973, 1986, 1996, 2014, and 2021 together with qualitative data were used. LULC classification scheme employed a supervised classification method with the application of the maximum likelihood algorithm technique. In the period between 1973 and 2021, agriculture, bare land, and settlement showed areal increment by 153.13%, 295.57%, and 49.03% with the corresponding increased annual rate of 1.93%, 2.86%, and 0.83%, respectively. On the contrary, forest, wood land, bushland, grass land, and water body decreased by 29.97%, 1.36%, 28.16%, 8.63%, and 84.36% during the study period, respectively. During the period, major LULC change dynamics were also observed; the majority of woodland was converted to agriculture (757.8 km2) and grassland (531.3 km2); and forests were converted to other LULC classes, namely woodland (766.5 km2), agriculture (706.1 km2), grassland (34.6 km2), bushland (31.9 km2), settlement (20.5 km2), and bare land (14.3 km2). LULC changes were caused by the expansion of agriculture, settlement, overgrazing, infrastructure development, and fire that were driven by population growth and climate change, and supplemented by inadequate policy and institutional factors. Social and environmental importance and values of land uses and land covers in the study area necessitate further assessment of potential natural resources' user groups and valuation of ecosystem services in the study area. Hence, we suggest the identification of potential natural resource-based user groups, and assessment of the influence of LULC changes on ecosystem services in Bale Mountains Eco Region (BMER) for the sustainable use and managements of land resources.

Comparative assessment of land use land cover at different temporal scales in Imphal-Iril river catchment, Manipur, India

ABSTRACT Land use land cover (LULC) assessment is essential for monitoring naturally available land resources through spatio-temporal changing pattern to protect land resources sustainably. The primary objective is to identify the changing pattern in LULC over last 10 years in Imphal-Iril catchment. Multi-spectral datasets of Landsat 7 and 8 were used for five different time periods (2011, 2013, 2015, 2017 and 2019). Supervised classification method was employed to generate the classified LULC maps using the maximum likelihood classifier approach in ArcGIS 10.3. In this assessment, five major land use classes have been classified as vegetation, agriculture, settlement, barren land and water body, and the dominant land cover is vegetation. The classified images obtained from this technique were checked for their accuracy using statistical kappa co-efficient and overall accuracy. According to this study, settlement and barren areas have been significantly increased by 72.25% (39.03 km2) and 22.60% (118.29 km2) during the last 10 years, whereas areas assigned to agriculture, vegetation and water bodies have decreased by 27.26% (72.16 km2), 7.59% (73.22 km2) and 60.39% (11.94 km2), respectively. LULC map for the year 2030 was also predicted using MOLUSCE model in QGIS. The results of this assessment will be significant to the policy implicates for sustainable LULC management in the region.

Water purification – an essential service from forest ecosystems, and farming practices in the Pong River Basin, Northeast Thailand

Deforestation often results in biodiversity depletion, land degradation, and loss of forest ecosystem services such as water purification. This research examined the relationship between natural forest areas in the Pong River Basin (PRB), northeast Thailand, and water quality in the Pong River system, one of the country's most polluted rivers. Water parameters were measured in 15 tributaries where samples were collected 500 m upstream from the outlets, and 2–5 m from the streambanks during summer, rainy, and winter seasons in 2021. Land use and land cover (LULC) were classified by a supervised classification method in QGIS using LANDSAT8 satellite imagery and USGS SRTM data. Vegetative surveys were conducted near the sampling locations which included species richness and density. Water characteristics at all the sites were under Thailand's Water Quality Standard, except a level of biological oxygen demand (avg. BOD = 3.52 mg/L). The upper PRB, covered with more forested areas with higher tree densities along streambanks (∼43 % of the basin, 619 trees/ha), revealed better water quality including a higher concentration of dissolved oxygen (avg. DO = 7.57 mg/L) and a lower level of fecal coliform bacteria (avg. FCB = 642 MPN/100 mL). In contrast, the lower PRB having larger areas of farmlands and smaller forests (∼63 % and 23 %, respectively) yielded higher concentrations of FCB (avg. = 1,353 MPN/100 mL) and phosphates (avg. PO43- = 0.34 mg/L). Positive correlations were found between the amount of forestlands (% of the sub-basin) with DO and nitrate concentrations (r = 0.468 and 0.409, p-value < 0.05, respectively), while agricultural areas were correlated with FCB (r = 0.414, p-value < 0.05). Amounts of forests determine water purity, while farmlands denoting dirtiness. These findings show that water purification is an essential forest ecosystem service, critical for stream health.

The examination of the relationship between land cover and land surface temperature in Karang Baru Sub-District, Aceh Province, Indonesia

Abstract Land cover change can lead to variations in land surface temperature, primarily because different materials have varying capacities to receive, absorb, and re-emit solar radiation. Karang Baru, situated as the capital of Aceh Tamiang Regency near North Sumatra, one of the most densely populated Regency, has been experiencing continuous development since 2002, following its separation from East Aceh Regency. The development in Karang Baru has resulted in a noticeable influence on land surface temperature. The primary objective of this study is to investigate the relationship between land cover and land surface temperature in Karang Baru from 2002 to 2022. This research classified land cover using the supervised classification method employing the Maximum Likelihood Algorithm, and land surface temperature was estimated using the Split Window Algorithm method. Subsequently, simple regression analysis in this context indicates a quantitative approach to understanding the relationship between land cover and land surface temperature in Karang Baru, providing valuable insights for environmental and urban planning considerations. The findings of this research indicate that the type of land cover indeed influences the land surface temperature in Karang Baru. In summary, this study establishes that land cover can substantially impact the regional surface temperature, with temperature increases associated with surfaces featuring buildings, roads, fields, or similar materials. In contrast, vegetated surfaces tend to exhibit lower land surface temperatures in the area.

K‐Fuse: Credit card fraud detection based on a classification method with a priori class partitioning and a novel feature selection strategy

AbstractOnline transactions have become the dominant and most popular form of online payment in today's digital economy. Due to the growing popularity of e‐commerce and the convenience it offers, both consumers and businesses are rapidly adopting online transactions. Notably, credit cards have become one of the most popular and standard online payment methods. However, it should be noted that credit card transactions are not without challenges. In particular, detecting and preventing fraudulent transactions is a major concern of the online payment system. It is difficult to find an effective detection model that can detect the new patterns created by fraudsters, due to the constant evolution of their methods to exploit the vulnerability of current security protocols. These fraud patterns are evolving and may not correspond to existing documented models, leading to a reduction in their identification. In addition, the customer's behavior can affect the model detection as it is susceptible to change based on factors such as economic conditions, trends, and individual circumstances. When consumers deviate from their typical behavior, the model may generate false alerts, thereby reducing its ability to differentiate between legitimate and fraudulent transactions. This article presents a new supervised detection model, called K‐Fuse, which introduces an unsupervised phase in order to detect fraud patterns that may correspond to innovative models introduced by fraudsters. K‐Fuse is a supervised classification method that fuses three steps consisting of (i) unsupervised clustering to identify hidden patterns of transactions in a dataset, (ii) a novel feature selection criterion based on the unsupervised results, and (iii) supervised classification to exploit the results of clustering and feature selection to predict new transactions as fraudulent or legitimate.

A Novel Classification Algorithm Based on the Synergy Between Dynamic Clustering with Adaptive Distances and K-Nearest Neighbors

This paper introduces a novel supervised classification method based on dynamic clustering (DC) and K-nearest neighbor (KNN) learning algorithms, denoted DC-KNN. The aim is to improve the accuracy of a classifier by using a DC method to discover the hidden patterns of the apriori groups of the training set. It provides a partitioning of each group into a predetermined number of subgroups. A new objective function is designed for the DC variant, based on a trade-off between the compactness and separation of all subgroups in the original groups. Moreover, the proposed DC method uses adaptive distances which assign a set of weights to the variables of each cluster, which depend on both their intra-cluster and inter-cluster structure. DC-KNN performs the minimization of a suitable objective function. Next, the KNN algorithm takes into account objects by assigning them to the label of subgroups. Furthermore, the classification step is performed according to two KNN competing algorithms. The proposed strategies have been evaluated using both synthetic data and widely used real datasets from public repositories. The achieved results have confirmed the effectiveness and robustness of the strategy in improving classification accuracy in comparison to alternative approaches.

Land Use Change Mapping and Analysis Using Remote Sensing and GIS: A Case Study in Tam Ky City, Quang Nam Province, Vietnam

Changes in land use/land cover (LULC) play a critical role in effective natural resource management, monitoring, and development, particularly within the realm of urban planning. In the examination of Tam Ky city, Quang Nam province, Vietnam, spanning from 2000 to 2020, remote sensing and Geographic Information System (GIS) techniques were employed. The Landsat satellite data (Landsat 7 ETM+ for 2000, Landsat 5 TM for 2010, and Landsat 8 OLI for 2022) underwent analysis using the supervised classification method in ArcGIS 10.8 software to identify and categorize six primary LULC classes: water bodies, agriculture, settlements, vegetation, construction, and bare soil/rocks. The reliability of the classification was evaluated through k values, revealing high accuracy with values of 0.951, 0.953, and 0.950 for the years 2000, 2010, and 2020, respectively. Notable shifts in LULC were observed during the period from 2000 to 2020. The areas covered by vegetation and settlements expanded by 53 and 1300 ha, respectively, while water bodies, agriculture, construction, and bare soil/rocks experienced reductions of 466, 48, 413, and 425 ha, respectively. To facilitate a rapid assessment, the study also incorporated the normalized difference vegetation index (NDVI) and normalized difference built-up index (NDBI). The trends identified in this study are consistently aligned with the results of the supervised classification. The identified changes in LULC pose a substantial environmental threat, and the study's outcomes serve as a valuable asset for future land use planning and management in the area. The method's high accuracy enhances the dependability of the results, making them crucial for well-informed decision-making and sustainable development initiatives.

Improvement of land cover classification accuracy by training sample clustering

The subject of this article is land cover classification based on geospatial data. The supervised classification methods are appropriate for most of the thematic tasks of remote sensing because they provide the opportunity to set the characteristics of the initial classes in the form of a training sample set, in contrast to unsupervised methods. There are many approaches to processing such a set; however, their common disadvantage is that they do not consider the factor of training sample separability. This characteristic indicates the extent to which signatures representing different classes do not overlap. A low degree of separability is inherent in high-level training sample mixing. Thus, separability affects classification accuracy. One possible ways to increase separability is training sample clustering. Considering the above, the goal of this study is to develop a training sample clustering technique to improve land cover classification accuracy by increasing the separability of training samples. The tasks of this work are as follows: 1) develop a method for training sample separability assessment; 2) develop a training sample clustering technique based on training sample separability; 3) test the effectiveness of the developed technique by applying it to experimental land cover classification. In the experiments, two land cover classifications were obtained for each of the two selected study areas (i.e., one before and another after training sample clustering. Six land cover classes were defined for each experiment. The training samples were selected for each class. Conclusions. After the application of the developed technique, an increase in the separability of the training samples was evidenced by the developed separability index. In turn, this approach led to an improvement in land cover classification. For the first experiment, this was evidenced by an increase in the overall accuracy and kappa coefficient by 20% (from 63 to 83%) and 21% (from 60% to 81%), respectively. In the second experiment, the increase was 4% (from 77% to 81%) and 5% (from 66% to 71%), respectively.

Determining future scenarios of urban areas with cellular automata/Markov Chain Model method; example of Ereğli District Konya-Türkiye (2030–2040)

As a result of the rapid increase in the world population, the earth surface has started to be damaged due to natural and artificial effects. The extent of the damage to nature can be determined by examining the temporal changes of land use and land cover (LULC). In order to offer healthier and more sustainable living spaces, scientists have produced many studies on the changes in nature. Within the scope of this study, 5 basic training classes were created with the help of Landsat satellite images and CORINE data, covering the period of 1985–2018 for Ereğli-Bor Sub-Basin, which is one of the 9 sub-basins of Konya Closed Basin located in the Central Anatolian Region of Türkiye. Landsat Satellite images, Google Earth Program and CORINE data were overlaid to create a basic training class as artificial areas, agricultural areas—pasture areas—forest areas and wetlands and these areas were classified by supervised classification method. The study was carried out on an area of approximately 331057 ha in and around Ereğli district. Modeling was carried out with the Cellular Automata (CA) Markov Chain Model to determine the urban development potential in the region. In order to estimate the modeling accuracy, the 2018 prediction model was created according to the 2018 reference map, and the validation between the two data was analyzed with the kappa statistics. According to kappa statistics values, it was determined that K_location and K_standard values were 0.9301 and 0.8935, respectively. As a result of the validation in sufficient standards, future prediction models were applied; future models and result maps were prepared for the years 2030–2040. According to the modeling results, it is estimated that the artificial area class in Ereğli district will reach 122.74 km2 by 2030 and 142.24 km2 in 2040. In addition, it was expressed in detail with the prediction results and maps that there will be a decrease in pasture, forest and agricultural areas in the region until 2030 and 2040. As a result, it is predicted that the ecological balance in the region will change and agricultural production may decrease as a result of the decline in agricultural pasture and forest areas. For this reason, it has been revealed that it is important for the future of humanity that plans such as environmental layout and master development plans to be made by regional manager in the region for the future should be planned in line with the results to be obtained as a result of future prediction models.

Multivariate Time Series Feature Extraction and Clustering Framework for Multi-Function Radar Work Mode Recognition

Multi-Function Radars (MFRs) are sophisticated sensors with great agility and flexibility in adapting their transmitted waveform and control parameters. The recognition of MFR work modes based on the intercepted pulse sequences plays an important role in interpreting the functional purpose and threats of a non-cooperative MFRs. However, due to the increased flexibility of MFRs, radar work modes with emerging new modulations and control parameters always appear, and the supervised classification method suffers performance degradation or even failure. Unsupervised learning and clustering of MFR pulse sequences becomes urgent and important. This paper establishes a unified multivariate MFR time series feature extraction and clustering framework for MFR work mode recognition. At first, various features are collected to form the feature set. The feature set includes features extracted through deep learning based on recurrent auto-encoders, multidimensional time series toolkit features, and manually crafted features for radar inter-pulse modulations. Subsequently, several feature selection algorithms, combined with different clustering and classification methods, are used for the selection of an “optimal” feature subset. Finally, the effectiveness and superiority of the proposed framework and selected features are validated through simulated and measured datasets. In the simulated dataset containing 20 classes of work modes, under the most severe non-ideal conditions, we achieve a clustering purity of 73.46% and an NMI of 84.28%. In the measured dataset with seven classes of work modes, we achieve a clustering purity of 86.96% and an NMI of 90.10%.

Crop type classification and spatial mapping in River Nile and Northern State, Sudan, using Sentinel-2 satellite data and field observation

Maintaining productive farmland necessitates precise crop mapping and identification. While satellite remote sensing makes it possible to generate such maps, there are still issues to resolve, such as how to choose input data and the best classifier algorithm, especially in areas with scarce field data. Accurate assessments of the land used for farming are a crucial part of national food supply and production accounting in many African countries, and to this end, remote sensing tools are being increasingly put to use. The aim of this study was to assess the potentiality of Sentinel-2 to distinguish and discriminate crop species in the study area and constraints on accurately mapping cropping patterns in the winter season in River Nile and Northern State, Sudan. The research utilized Sentinel-2 Normalized Different Vegetation Index (NDVI) at 10 m resolution, unsupervised and supervised classification method with ground sample and accuracy assessment. The results of the study found that the signatures of grain sorghum, wheat, okra, Vicia faba, alfalfa, corn, haricot, onion, potato, tomato, lupine, tree cover, and garlic have clear distinctions, permitting an overall accuracy of 87.38%, with trees cover, onion, wheat, potato, garlic, alfalfa, tomato, lupine and Vicia faba achieving more than 87% accuracy. Major mislabeling problems occurred primarily in irrigated areas for grain sorghum, okra, corn, and haricot, in wooded areas comprised of small parcels of land. The research found that high-resolution temporal images combined with ground data had potential and utility for mapping cropland at the field scale in the winter.

Galaxy assembly revealed by globular clusters

Many observable properties of globular clusters (GCs) provide valuable insights for unveiling the hierarchical assembly of their host galaxy. For the Milky Way (MW) in particular, GCs from different accreted satellite galaxies show distinct chemical, spatial, kinematic, and age distributions. Here we examine such clustering features for model GC populations in simulated galaxies, which are carefully selected to match various observational constraints of the MW assembly. We evaluate several widely used clustering, dimensionality reduction, and supervised classification methods on these model GCs, using 10 properties that are observable in the MW. We can categorize in-situ and ex-situ formed GCs with about 90% accuracy, based solely on their clustering features in these 10 variables. The methods are also effective in distinguishing the last major merger in MW analogs with similar accuracy. Although challenging, we still find it possible to identify one, and only one, additional smaller satellite. We develop a new technique to classify the progenitors of MW GCs by combining several methods and weighting them by the validated accuracy. According to this technique, about 60% of GCs belong to the in-situ group, 20% are associated with the Gaia-Sausage/Enceladus event, and 10% are associated with the Sagittarius dwarf galaxy. The remaining 10% of GCs cannot be reliably associated with any single accretion event.