Landscape Classification Research Articles

The feasibility of using national‐scale datasets for classifying wetlands in Arizona with machine learning

AbstractThe advent of machine learning techniques has led to a proliferation of landscape classification products. These approaches can fill gaps in wetland inventories across the United States (U.S.) provided that large reference datasets are available to develop accurate models. In this study, we tested the feasibility of expediting the classification process by sourcing requisite training and testing data from existing national‐scale land cover maps instead of customized sample sets. We created a single map of water and wetland presence by intersecting water and wetland classes from available land cover products (National Wetland Inventory, Gap Analysis Project, National Land Cover Database and Dynamic Surface Water Extent) across the U.S. state of Arizona, which has fewer wetland‐specific mapping products than other parts of the U.S. We derived classified samples for four wetland classes from the combined map: open water, herbaceous wetlands, wooded wetlands and non‐wetland cover. In Google Earth Engine, we developed a random forest model that combined the training data with spatial predictor variables, including vegetation greenness indices, wetness indices, seasonal index variation, topographic parameters and vegetation height metrics. Results show that the final model separates the four classes with an overall accuracy of 86.2%. The accuracy suggests that existing datasets can be effectively used to compile machine learning training samples to map wetlands in arid landscapes in the U.S. These methods hold promise for the generation of wetland inventories at more frequent intervals, which could allow more nuanced investigations of wetland change over time in response to anthropogenic and climatic drivers.

The Early Formation of the Moscow School of Landscape Painting. Landscape Class of the Moscow School of Painting and Sculpture (MUZHV), 1844–1857

The article delves into the formative years of the landscape class at the Moscow School of Painting and Sculpture (MUZHV), whose growth parallels the advancement of the landscape genre in Russia during the latter half of the 1 9th century. It explores the pioneering efforts of Karl Ivanovich Rabus, the inaugural head of the class, and evaluates his instructional techniques in landscape painting. His methods introduced academic traditions of landscape painting to the School, thereby laying the groundwork for its future development. Additionally, the article delineates the challenges encountered by the nascent landscape painting tradition, attributable to a range of internal and external factors.Drawing from archival records and periodical sources, the author endeavors to reconstruct a detailed portrayal of the class's early years (1844-1857), emphasizing the pivotal role of students in the evolution of artistic life and the formation of the art market in Moscow. The research enhances scholarly knowledge by bringing into focus the identities of the class's students, most of whom are lesser-known artists, alongside their artistic creations, enriching our comprehension of Russian landscape painting during the late 1800s.

On the effectiveness of multi-scale landscape metrics in soil organic carbon mapping

Soil organic carbon (SOC) mapping delivers invaluable information to the global carbon budget and climate change mitigation endeavour. Environmental variables at sample locations are frequently used as explanatory variables for simulating the spatial distribution of soil properties. However, these may not fully capture the spatial information generated by soil-forming processes. We applied multi-scale landscape metrics that can comprehensively characterize the surrounding landscape information of sample points. The metrics were extracted at two levels (landscape level and class level) and include the diversity, shape and area, fragmentation and connectivity with a buffer distance from 500 m to 5,000 m. We then investigated its effectiveness as environmental variables via recursive feature elimination, random forests, and quantile regression forests. The Jianghan Plain, China, was selected as the study area, where over 19,000 topsoil samples were collected. Results indicated that multi-scale landscape metrics enhanced the predictability of SOC mapping, with R2 increased by 43 %. Specifically, Shannon’s diversity index, the percentage of landscape index, interspersion and juxtaposition index, and patch cohesion index outperformed environmental variables that were extracted at the sample location. In addition, the relationships between SOC and landscape metrics were found to be scale-dependent. Landscape metrics demonstrated significant explanatory capacity for SOC across various spatial scales. Notably, as the scale surpassed 3,000 m, there was a discernible improvement in the explanatory effectiveness of the landscape metrics for SOC. Our findings highlight that landscape metrics are effective in characterising the soil-landscape relationship that is generated by multi-scale natural and anthropogenic soil-forming processes. Meanwhile, knowledge of the intricate relationship between landscape characteristics and SOC is crucial for informing land management decisions aimed at enhancing carbon sequestration, mitigating climate change, and maintaining soil fertility.

Determination of the Change of Agricultural Landscapes Based on CORINE Land Cover Agricultural Land Classes Using GIS and Visual Quality Value with the AHP Method: The Case of Tekı̇rdağ Province

Agricultural landscapes contribute positively to the visual texture of the city and change over time. In this context, this study aims to determine the visual landscape quality of agricultural landscapes based on parameters on expert approach the hypothesis; it is important and necessary to determine the visual landscape values based on the change in agricultural landscapes over time. Within the framework of the developed hypothesis, it aims to make an expert-based visual landscape assessment of the agricultural landscapes of Tekirdağ Province, located in the northwestern part of Türkiye, based on nine subclasses created according to the CORINE land cover (CLC) classes, on the basis of 4 main parameters. Using the AHP technique, the priorities of the parameters and agricultural landscape classes and the relationships between basic parameters and visual preferences were determined. In this direction, the questions; What are the priorities of parameters that are effective in determining the visual quality of agricultural landscapes on the basis of expert approach?; What kind of changes have occurred in agricultural landscapes during the years 1990-2000-2006-2012-2018 in 9 subclasses created according to the CLC classes? and; How should the visual landscape quality values of the agricultural landscape subclasses created according to the CLC classes be ranked? The results of the study can be used as a tool in landscape planning and management studies as a factor in strengthening landscape quality.

The impact of the urban landscape on PM2.5 from LCZ perspective: A case study of Shenyang

Based on the World City Database Access Portal tool (WUDAPT), the local Climate zoning (LCZ) classification of Shenyang(China) was carried out. With 8 environmental monitoring stations at the observation of the study area, LCZ landscape categories in the buffer zone of 500 m, 1000 m, and 3000 m were analyzed using ArcMap and Fragstats. The sensitivity analysis of the landscape category index correlated with PM2.5 was conducted to select the representativeness index. Matlab was used to construct the regression equation between the LCZ categories and PM2.5. The results show that AI_500, LSI_500, PD_500, LSI_1000, AI_1000, LSI_3000, and AI_3000 can be used as representative landscape indices to represent the aggregation and diversity of LCZ categories. The regression equations for PM2.5 in all four seasons include factors influencing the LCZ category landscape. This study analyzed the relationship between PM2.5 distribution and urban patterns from the perspective of the LCZ types. It provided the scientific reference for rational planning of urban layout and coordination of environment and urban development.

Morphology's importance for farmland landscape pattern assessment and optimization: A case study of Jiangsu, China

Farmland landscape pattern (FLP) is critical in ensuring sustainable agricultural production. Amidst the challenge of increasingly fragmented FLP, accurate assessment and optimization strategies are essential. Traditional FLP assessment studies lack significance in ecological representation because they ignore landscape types and their characteristics and overlook changes in landscape details based on overall results. These factors lead to uncertain assessment outcomes. This study introduces a farmland landscape classification system based on morphological spatial pattern analysis to address this. Special attention is given to the internal core and external edge farmland landscape types, resulting in an innovative FLP assessment system being developed. The results of the Jiangsu Province reveal significant patchiness and marginalization of core farmland, particularly worsening from north to south. While recent trends show a slight improvement in the overall situation and effective perforation suppression, challenges persist with low-value FLP expansions in each agricultural zone. This study emphasizes integrating external edge farmland and overall pattern protection for optimal FLP. The improved assessment method identifies areas for scale improvement and fragmentation risks, offering valuable insights for permanent basic farmland policy. Overall, this study enriches the analytical framework of FLP from a morphological perspective, aiding in precise assessment and tailored optimization strategies.

Functional trait mismatch between native and introduced bee pollinators servicing a global fruit crop

BackgroundUnderstanding connections between biodiversity and ecosystem services can be enhanced by shifting focus from species richness to functional trait-based approaches, that when paired with comparative phylogenetic methods can provide even deeper insights. We investigated the functional ecology and phylogenetic diversity of pollination services provided by hymenopteran insects visiting apple flowers in orchards surrounded by either ‘natural’ or ‘disturbed’ landscapes in New South Wales, Australia. We assessed whether morphological and behavioural traits (hairiness, body size, glossa length, pollen load purity, and probability of loose pollen) exhibited non-random phylogenetic patterns. Then, explored whether bees, the primary pollinators in this system, filled unique or overlapping functional entities (FEs). For each landscape, we calculated phylogenetic diversity and used FEs to assess functional richness, evenness, and diversion.ResultsA phylogenomic matrix based on ultraconserved elements (UCEs; 1,382,620 bp from 1,969 loci) was used to infer a fully-resolved and well-supported maximum likelihood phylogeny for 48 hymenopteran morphospecies. There was no significant difference in species richness between landscape categories. Pollinator communities at natural sites had higher phylogenetic complexity (X = 2.37) and functional divergence (x̄ = 0.74 ± 0.02 s.e.) than disturbed sites (X = 1.65 and x̄ = 0.6 ± 0.01 s.e.). Hairiness showed significant phylogenetic clustering (K = 0.94), whereas body size, glossa length, and loose pollen showed weaker non-random phylogenetic patterns (K between 0.3–0.5). Pollen load purity showed no association with phylogeny. The assemblage of 17 bee morphospecies comprised nine FEs: eight FEs consisted of native bees with three containing 65% of all native bee taxa. The introduced honey bee (Apis mellifera) occupied a unique FE, likely due to its different evolutionary history. Both landscape types supported six FEs each with three overlapping: two native bee FEs and the honey bee FE.ConclusionsBee hairiness was the only functional trait to exhibit demonstrable phylogenetic signal. Despite differences in species richness, and functional and phylogenetic diversity between orchard landscape types, both maintained equal bee FE numbers. While no native bee taxon was analogous to the honey bee FE, four native bee FEs shared the same hairiness level as honey bees. Health threats to honey bee populations in Australia will likely disrupt pollination services to apple, and other pollination-dependent food crops, given the low level of functional redundancy within the investigated pollinator assemblages.

A hierarchical method and workflow for the semi‐automated mapping of valley bottom geomorphic units using publicly available remote sensing datasets

AbstractGeomorphic units (GUs) are the landforms that make up the valley bottom and are produced by fluvial processes that determine river structure and function. Mapping of GUs can be used to interpret river type and behaviour and to analyse river condition and recovery processes. The advancement of remote sensing technologies and big‐data acquisition are enabling the development and operationalisation of tools to semi‐automate the mapping of assemblages of GUs across large spatial areas. In this study, we develop a hierarchical method that combines a landscape classification approach (Geomorphons) with supervised classification using light detection and ranging data (LiDAR) and satellite images to semi‐automate the mapping of GUs across valley bottoms. We have also produced a new method for identifying and mapping pools in the absence of bathymetry data. We applied our method on 78 river sections in coastal catchments of NSW, Australia. We were able to identify 20 refined GU types and four further sub‐types of bank‐attached bars. Our method produced GU maps that are consistent with desktop manual delineation from aerial images and digital elevation models. Our hierarchical method offers GU maps with varying accuracy and resolution, accommodating a user's decisions regarding amount of effort invested relative to map quality and accuracy required. Initial runs produce maps with 12 preliminary GUs with 61%–75% consistency when compared to desktop manual mapping. With additional effort and manual corrections, a higher level of GU identification is possible (i.e. refined GU mapping increases the consistency to 70%–81%). The delineation of more intricate sub‐types of GU or sub‐units on compound GUs, which is essential for interpretation of river behaviour, condition, and recovery, still requires on‐site field verification to achieve the best results.

A Stochastic Landscape Approach for Protein Folding State Classification.

Protein folding is a critical process that determines the functional state of proteins. Proper folding is essential for proteins to acquire their functional three-dimensional structures and execute their biological role, whereas misfolded proteins can lead to various diseases, including neurodegenerative disorders like Alzheimer's and Parkinson's. Therefore, a deeper understanding of protein folding is vital for understanding disease mechanisms and developing therapeutic strategies. This study introduces the Stochastic Landscape Classification (SLC), an innovative, automated, nonlearning algorithm that quantitatively analyzes protein folding dynamics. Focusing on collective variables (CVs) - low-dimensional representations of complex dynamical systems like molecular dynamics (MD) of macromolecules - the SLC approach segments the CVs into distinct macrostates, revealing the protein folding pathway explored by MD simulations. The segmentation is achieved by analyzing changes in CV trends and clustering these segments using a standard density-based spatial clustering of applications with noise (DBSCAN) scheme. Applied to the MD-based CV trajectories of Chignolin and Trp-Cage proteins, the SLC demonstrates apposite accuracy, validated by comparing standard classification metrics against ground-truth data. These metrics affirm the efficacy of the SLC in capturing intricate protein dynamics and offer a method to evaluate and select the most informative CVs. The practical application of this technique lies in its ability to provide a detailed, quantitative description of protein folding processes, with significant implications for understanding and manipulating protein behavior in industrial and pharmaceutical contexts.

The clock gene BHLHE40 and atypical CCNG2 control androgen-induced cellular senescence as a novel tumor suppressive pathway in prostate cancer

BackgroundThe androgen receptor (AR) is a drug target used to inhibit AR and prostate cancer (PCa) growth. Surprisingly, treatment with supraphysiological androgen level (SAL), used in bipolar androgen therapy, inhibits growth of PCa suggesting a tumor-suppressive activity by SAL. SAL was shown to induce cellular senescence in PCa.MethodsRNA-seq and transcriptome analysis, ChIP-seq, human 3D PCa spheroids, mouse xenografted castration-resistant PCa, knockdown and overexpression, Co-immunoprecipitation (Co-IP), translocation analysis, immune detection, qRT-PCR, protein–protein interaction modelling.ResultsHere, mice xenografts with castration-resistant PCa tumors show that SAL inhibits cancer growth in vivo suggesting that SAL activates a tumor-suppressive mechanism. RNA-seq and ChIP-seq revealed the clock gene BHLHE40 is a novel direct AR target. Compared to adjacent human prostate tissues, the expression of BHLHE40 is reduced in PCa tumors and associated with reduced survival. Knockdown suggests that BHLHE40 mediates SAL-induced cellular senescence including tumor spheroids. Interestingly, a large overlap of differentially expressed gene sets was identified between BHLHE40 and SAL leading to the identification of four classes of SAL-BHLHE40 transcriptome landscapes. Co-IP and modelling suggest binding of BHLHE40 to AR and their co-translocation into nucleus by SAL treatment. Further, RNA-seq and ChIP-seq analysis indicate that the atypical tumor suppressive cyclin G2 emerged as a novel downstream target of BHLHE40 and a mediator of SAL-induced cellular senescence.ConclusionsThe data provide evidence of the tumor suppressive activity of SAL and a novel signaling by the AR-BHLHE40-CCNG2 axis for androgen-induced cellular senescence, linking circadian rhythm factor to androgen signaling as a novel tumor suppressive pathway.

SLOW TOURISM AS A CONTEMPORARY TENDENCY IN THE TOURISM MARKET: IMAGE CONTENT AND GEOTAG ANALYSIS ON INSTAGRAM

Although it originated from the Slow Food and Cittaslow movements, slow tourism has spread to many spheres of tourism. The purpose of the paper is to investigate what people associate with slow tourism, and how and where they engage in slow tourism. Social media content analysis on Instagram, using hashtag #slowtourism, was performed in the research. The general conclusion is that people mostly associate slow tourism with different forms of architecture and landscape, and to lesser extent with transport and food. Also, there was significant content variation within the architecture and landscape categories, suggesting that people associate different types of architecture and tourism destinations with slow tourism. When it comes to location, the majority of photos were taken in Italy, which is not surprising, considering that Slow Food and Cittaslow movements originated in this country.

Evaluating Machine-Learning Algorithms for Mapping LULC of the uMngeni Catchment Area, KwaZulu-Natal

Analysis of land use/land cover (LULC) in catchment areas is the first action toward safeguarding freshwater resources. LULC information in the watershed has gained popularity in the natural science field as it helps water resource managers and environmental health specialists develop natural resource conservation strategies based on available quantitative information. Thus, remote sensing is the cornerstone in addressing environmental-related issues at the catchment level. In this study, the performance of four machine learning algorithms (MLAs), namely Random Forests (RFs), Support Vector Machines (SVMs), Artificial Neural Networks (ANNs), and Naïve Bayes (NB), were investigated to classify the catchment into nine relevant classes of the undulating watershed landscape using Landsat 8 Operational Land Imager (L8-OLI) imagery. The assessment of the MLAs was based on a visual inspection of the analyst and commonly used assessment metrics, such as user’s accuracy (UA), producers’ accuracy (PA), overall accuracy (OA), and the kappa coefficient. The MLAs produced good results, where RF (OA = 97.02%, Kappa = 0.96), SVM (OA = 89.74%, Kappa = 0.88), ANN (OA = 87%, Kappa = 0.86), and NB (OA = 68.64%, Kappa = 0.58). The results show the outstanding performance of the RF model over SVM and ANN with a significant margin. While NB yielded satisfactory results, its sensitivity to limited training samples could primarily influence these results. In contrast, the robust performance of RF could be due to an ability to classify high-dimensional data with limited training data.

Fashion Design Classification Based on Machine Learning and Deep Learning Algorithms: A Review

Integration of machine learning algorithms in fashion design classification brought transformational change by allowing the automated analysis, categorization, and prediction of fashion items based on different attributes of the item. This paper reviews the state-of-the-art in fashion design classification through machine learning techniques. Review of literature, methodology, and challenges in this area indicate that an array of algorithms and methods, stretching from traditional machine learning algorithms to convolutional neural networks and further to transfer learning approaches, is being tried and tested. In this paper, I will discuss performance comparison among several machine learning algorithms, pinning their strengths, limitations, and possible applications in the fashion industry. We further elaborate on crucial challenges, such as touching on the issue of data variability, interpretability, and others on ethical consideration issues, all pointing to the need for fairness and sustainability with respect to the representation of reality in algorithmic decision-making. This paper aims to inform researchers, practitioners, and stakeholders of the opportunities and challenges brought about by the use of machine learning in a fast-paced world like fashion, hence demystifying current directions in landscape classification of fashion design.

Confinement energy landscape classification reveals membrane receptor nano-organization mechanisms

The cell membrane organization has an essential functional role through the control of membrane receptor confinement in micro- or nanodomains. Several mechanisms have been proposed to account for these properties, although some features have remained controversial, notably the nature, size, and stability of cholesterol- and sphingolipid-rich domains or lipid rafts. Here, we probed the effective energy landscape acting on single-nanoparticle-labeled membrane receptors confined in raft nanodomains— epidermal growth factor receptor (EGFR), Clostridium perfringens ε-toxin receptor (CPεTR), and Clostridium septicum α-toxin receptor (CSαTR)—and compared it with hop-diffusing transferrin receptors. By establishing a new analysis pipeline combining Bayesian inference, decision trees, and clustering approaches, we systematically classified single-protein trajectories according to the type of effective confining energy landscape. This revealed the existence of only two distinct organization modalities: confinement in a quadratic energy landscape for EGFR, CPεTR, and CSαTR (A), and free diffusion in confinement domains resulting from the steric hindrance due to F-actin barriers for transferrin receptor (B).The further characterization of effective confinement energy landscapes by Bayesian inference revealed the role of interactions with the domain environment in cholesterol- and sphingolipid-rich domains with (EGFR) or without (CPεTR and CSαTR) interactions with F-actin to regulate the confinement energy depth. These two distinct mechanisms result in the same organization type (A). We revealed that the apparent domain sizes for these receptor trajectories resulted from Brownian exploration of the energy landscape in a steady-state-like regime at a common effective temperature, independently of the underlying molecular mechanisms. These results highlight that confinement domains may be adequately described as interaction hotspots rather than rafts with abrupt domain boundaries. Altogether, these results support a new model for functional receptor confinement in membrane nanodomains and pave the way to the constitution of an atlas of membrane protein organization.

Correction: Landscape classification with self-organizing map using user participation and environmental data: the case of the Seoul Metropolitan Area

Correction: Landscape classification with self-organizing map using user participation and environmental data: the case of the Seoul Metropolitan Area

Landscape classification with self-organizing map using user participation and environmental data: the case of the Seoul Metropolitan Area

This study aimed to develop a method for assessing landscapes using environmental data and user-generated data, which are commonly employed in landscape research. It focused on the Seoul metropolitan area in South Korea, devising evaluation indicators for five key concepts: naturalness, diversity, imageability, historicity, and disturbance. These indicators were used to assess the landscapes based on each index. We employed a self-organizing map, an artificial neural network technique, to categorize the landscape units and developed eight evaluation indicators for the five key concepts, organizing the study area’s landscapes into six distinct landscape units. This study identified landscape unit types with increased vulnerability to visual blight or heightened public awareness by considering both user characteristics and environmental attributes in the metropolitan area landscapes. Finally, we discussed future tasks for appropriate landscape management based on each landscape area’s characteristics to maintain and enhance landscape quality.

Analysis of landscape pattern vulnerability in Dasi river basin at the optimal scale

Since the reform and opening up in 1978, the Dasi River Basin within Jinan’s startup area from replacing old growth drivers with new ones (startup area) has experienced rapid urbanization and industrialization, and the landscape pattern has changed significantly, resulting in a series of eco-environmental problems. In order to more accurately identify the vulnerable areas of landscape pattern, understand their cause mechanism and changing laws, and provide a theoretical basis for the implementation of sustainable landscape pattern planning and management in the region. Four Landsat images of 2002, 2009, 2015 and 2020 were taken as data sources, and the optimal granularity of landscape pattern analysis was determined from the perspective of landscape level and class level by using the coefficient of variation method, granularity effect curve and information loss model, and the optimal amplitude was determined by using the grid method and semi-variance function. Then, the landscape vulnerability assessment model was constructed based on the optimal scale, and its spatiotemporal evolution characteristics and spatial autocorrelation were analyzed. The result showed that: (1) The optimal granularity of landscape pattern analysis in this study area was 80 m, and the optimal amplitude was 350 × 350 m. (2) During 2002–2020, the overall vulnerability of landscape pattern in the southern part of the study area showed an increasing trend, while that in the middle and northern parts showed a decreasing trend. (3) The mean values of the vulnerability index of the overall landscape pattern in 2002, 2009, 2015 and 2020 were 0.1479, 0.1483, 0.1562 and 0.1625, respectively, showing an increasing trend year by year. In terms of land use, during 2002–2020, the average vulnerability indices of forestland and built up land increased by 23.18% and 21.43%, respectively, followed by water body and bare land, increased by 12.18% and 9.52%, respectively, while the changes of cropland and grassland were relatively small, increasing by 5.36% and 5.65%, respectively. (4) During 2002–2020, the landscape pattern vulnerability showed a significant spatial positive correlation in terms of spatial distribution. The Low-Low areas were generally transferred from the southeastern and midwestern to the middle and northern, and the High–High areas were mainly transferred from the middle to the southern. Overall, the degree of the spatial agglomeration of the landscape pattern vulnerability showed an increasing trend.

A generalised ecohydrological landscape classification for assessing ecosystem risk in Australia due to an altering water regime

Abstract. Describing and classifying a landscape for environmental impact and risk assessment purposes is a non-trivial challenge because this requires region-specific landscape classifications that cater for region-specific impacts. Assessing impacts on ecosystems from the extraction of water resources across large regions requires a causal link between landscape features and their water requirements. We present the rationale and implementation of an ecohydrological classification for regions where coal mine and coal seam gas developments may impact on water. Our classification provides the essential framework for modelling the potential impact of hydrological changes from future coal resource developments at the landscape level. We develop an attribute-based system that provides representations of the ecohydrological entities and their connection to landscape features and make use of existing broad-level classification schemes into an attribute-based system. We incorporate a rule set with prioritisation, which underpins risk modelling and makes the scheme resource efficient, where spatial landscape or ecosystem classification schemes, developed for other purposes, already exist. A consistent rule set and conceptualised landscape processes and functions allow for the combination of diverse data with existing classification schemes. This makes the classification transparent, repeatable and adjustable, should new data become available. We apply the approach in three geographically different regions, with widely disparate information sources, for the classification, and provide a detailed example of its application. We propose that it is widely applicable around the world for linking ecohydrology to environmental impacts.

Calabi-Yau four-, five-, sixfolds as Pwn hypersurfaces: Machine learning, approximation, and generation

Calabi-Yau fourfolds may be constructed as hypersurfaces in weighted projective spaces of complex dimension five defined via weight systems of six weights. In this work, neural networks were implemented to learn the Calabi-Yau Hodge numbers from the weight systems, where gradient saliency and symbolic regression then inspired a truncation of the Landau-Ginzburg model formula for the Hodge numbers of any dimensional Calabi-Yau constructed in this way. The approximation always provides a tight lower bound, is shown to be dramatically quicker to compute (with computation times reduced by up to 4 orders of magnitude), and gives remarkably accurate results for systems with large weights. Additionally, complementary datasets of weight systems satisfying the necessary but insufficient conditions for transversality were constructed, including considerations of the interior point, reflexivity, and intradivisibility properties, overall producing a classification of this weight system landscape, further confirmed with machine learning methods. Using the knowledge of this classification and the properties of the presented approximation, a novel dataset of transverse weight systems consisting of seven weights was generated for a sum of weights ≤200, producing a new database of Calabi-Yau fivefolds, with their respective topological properties computed. Furthermore, an equivalent database of candidate Calabi-Yau sixfolds was generated with approximated Hodge numbers. Published by the American Physical Society 2024

An experimental comparison of pixel-based and object-based classifications with different machine learning algorithms in landscape pattern analysis – Case study from Quang Ngai city, Vietnam

In landscape pattern analysis, the choice of an efficient method for image classification is widely studied, but the features are mostly extracted from digital values by the traditional approach (Pixel-based) and modern approach (Object-based). In this study, we compared the performance of two supervised classification algorithms (Maximum likelihood classifier - MLC and Support vector machines - SVM). We used SPOT-5 image data from 2011 to analyze the landscape pattern of a complex territory in Quang Ngai City, Vietnam. We collected 215 ground-truth samples and classified them into seven landscape classes. The results showed that the overall accuracy of the classification of Pixel-based (MLC), Object-based (MLC), Pixel-based (SVM) and Object-based (SVM) was 67.9%, 74.0%, 72.1%, and 82.3%, respectively. The combination of the object-based approach and SVM algorithm had the best classification result, which reflected the current spatial distribution of land cover types accurately. In the next step, we computed landscape metrics from detailed images to compare quantitative parameters with the actual verification data sources. From these metrics results, we discussed how classification methods could affect landscape structure, and possible ways to improve the accuracy of landscape-pattern identification.