Search

Byblos, designated as a UNESCO World Heritage site, stands as one of Lebanon's most ancient urban centers, known for its expansive green spaces. However, ongoing urbanization threatens these valuable areas. This study uses a multi-faceted approach to evaluate the structure and landscape attributes of Byblos' urban and peri-urban forests (UPFs). Landscape canopy cover, diversity indices, forest structure, and a silhouette perceptual test were assessed across 24 streets in the city center, residential zones, and areas with heavy vehicular traffic. Findings reveal that 28% of Byblos' canopy cover is concentrated mostly in the northeastern region. Native tree species account for 30% of the total, and a notable variation in tree diversity exists among different land-use types (Shannon diversity index (H) was 1.02 for the city center, 1.35 for residential streets, and 0.64 for vehicular areas). Additionally, a normal J-shaped distribution of tree diameters was identified across all street types. This study highlights a correlation between tree silhouettes and visual preferences, with densely spreading canopies being favored. Residential trees demonstrate the highest structural diversity and varied blossoming seasons. This research represents the first investigation into the current state of urban forestry in Byblos and offers recommendations for sustainable management and planning strategies. [ABSTRACT FROM AUTHOR]

Warts are benign tumors infected by the Human Papillomavirus. Physicians and medical practitioners are endeavoring to identify the best wart treatment method. The present study finds the response of well-known wart treatment methods, namely immunotherapy and cryotherapy, towards the removal of predominant wart types such as plantar and common warts. The present study utilized the optimal feature space generated by the measure of Fisher score, information gain, and univariate statistical test. In addition, the proposed method finds the customized cost in terms of class weighted and non-class weighted to reduce the miss-classified instances for the positive class sample. The class-weighted and non-class-weighted approaches explicitly incorporated with the well-known classification algorithm extreme gradient boosting approach, which provides a maximum measure of true positive rate, true negative rate, positive predicted value, F-measure, and area under receiver operating characteristic curve of 100.00, 100.00, 100.00, 80.00, and 82.00% respectively on immunotherapy dataset, 100.00, 100.00, 100.00, 100.00, 92.00% respectively on cryotherapy dataset. While validating the performance on the benchmark dataset with the state-of-the-art approach, the proposed model gives an improvement of 6.40% to a maximum of 43.00% in terms of specificity on the immunotherapy dataset. However, the proposed model improves 3.33 - 30%, 5.70 -30%, and 0 - 31% in terms of accuracy, sensitivity, and specificity, respectively on cryotherapy dataset. Also, the proposed framework achieved a maximum sensitivity of 91.30%, which dominates the existing state-of-the-art approaches by a margin of 1.87% and 10.82%, respectively on the merged dataset. [ABSTRACT FROM AUTHOR]

A homogeneous distribution of students in a class is accepted as a key factor for overall success in primary education. A class of students with similar attributes normally increases academic success. It is also a fact that general academic success might be lower in some classes where students have different intelligence and academic levels. In this study, a class distribution model is proposed by using some data science algorithms over a small number of students' dataset. With unsupervised and semi-supervised learning methods in machine learning and data mining, a group of students is equally distributed to classes, taking into account some criteria. This model divides a group of students into clusters by the considering students' different qualitative and quantitative characteristics. A draft study is carried out by predicting the effectiveness and efficiency of the presented approaches. In addition, some process elements such as quantitative and qualitative characteristics of a student, data acquisition style, digitalization of attributes, and creating a future prediction are also included in this study. Satisfactory and promising experimental results are received using a set of algorithms over collected datasets for classroom scenarios. As expected, a clear and concrete evaluation between balanced and unbalanced class distributions cannot be performed since these two scenarios for the class distributions cannot be applicable at the same time. [ABSTRACT FROM AUTHOR]

Plant diseases may cause large yield losses, endangering both the stability of the economy and the supply of food. Convolutional Neural Networks (CNNs), in particular, are deep neural networks that have shown remarkable effectiveness in completing image categorization tasks, often outperforming human ability. It has numerous applications in voice processing, picture and video processing, and natural language processing (NLP). It has also grown into a centre for research on plant protection in agriculture, including the assessment of pest ranges and the diagnosis of plant diseases. In two plant phenotyping tasks, the function of a CNN (Convolutional Neural Networks) structure based on Residual Networks (ResNet) is investigated in this study. The majority of current studies on Species Recognition (SR) and plant infection detection have used balanced datasets for accuracy and experimentation as the evaluation criteria. This study, however, made use of an unbalanced dataset with an uneven number of pictures, organised the data into several test cases and classes, conducted data augmentation to improve accuracy, and-- most importantly--used multiclass classifier assessment settings that were helpful for an asymmetric class distribution. Furthermore with all these frequent issues, the paper addresses selecting the size of the data collection, classifier depth, necessary training time, and assessing the efficacy of the classifier when using various test scenarios. The Species Recognising (SR) and Identifying of Health and Infection Leaves (IHIL) tasks in this study have shown substantial improvement in performance for the ResNet 20 (V2) architecture, with Precision of 91.84% & 84.00%, Recall of 91.67% and 83.14%, and F1 scores of 91.49% & 83.19%, respectively. [ABSTRACT FROM AUTHOR]

Byblos, designated as a UNESCO World Heritage site, stands as one of Lebanon’s most ancient urban centers, known for its expansive green spaces. However, ongoing urbanization threatens these valuable areas. This study uses a multi-faceted approach to evaluate the structure and landscape attributes of Byblos’ urban and peri-urban forests (UPFs). Landscape canopy cover, diversity indices, forest structure, and a silhouette perceptual test were assessed across 24 streets in the city center, residential zones, and areas with heavy vehicular traffic. Findings reveal that 28% of Byblos’ canopy cover is concentrated mostly in the northeastern region. Native tree species account for 30% of the total, and a notable variation in tree diversity exists among different land-use types (Shannon diversity index (H) was 1.02 for the city center, 1.35 for residential streets, and 0.64 for vehicular areas). Additionally, a normal J-shaped distribution of tree diameters was identified across all street types. This study highlights a correlation between tree silhouettes and visual preferences, with densely spreading canopies being favored. Residential trees demonstrate the highest structural diversity and varied blossoming seasons. This research represents the first investigation into the current state of urban forestry in Byblos and offers recommendations for sustainable management and planning strategies.

Data selection is commonly used to reduce costs and energy usage by training on a subset of available data. However, determining the appropriate subset size requires extensive dataset knowledge and experimentation, limiting transferability. Varying the validation set also produces unstable results and wastes computational resources. In this paper, we propose a data selection method for dynamically determining subset ratios based on model performance using only a training set. The data search space is narrowed through weighted sampling, leveraging statistical selection patterns. Parallel analysis of class distributions identifies the most representative samples with high selection potential. Extensive experiments validate our approach and demonstrate improved training efficiency. Our method speeds up various subset ratios by up to 2.2x on CIFAR-10, 1.9x on CIFAR-100, 2.0x on TinyImageNet, and 2.3x on ImageNet with negligible accuracy drops. [Display omitted] Main flowchart of Dyna m ic Data S u bset S election for I mage C lassification (Music) to speed up neural network training, where reinforced data selection is performed every L epochs. Music is compatible with any data selection algorithm, and model parameters are updated using selected data subsets. The block diagram shows interactions between modules in the reinforced selection component. • Propose a data selection method to dynamically determine the subset ratio. • Study the role of intelligent data subset ratios. • Orthogonalize existing subset selection methods. • Two novel modules are used to narrow the data search. • Enable more efficient and sustainable AI development. [ABSTRACT FROM AUTHOR]

Data are being generated and used to support all aspects of healthcare provision, from policy formation to the delivery of primary care services. Particularly, with the change of emphasis from curative to preventive medicine, the importance of data-based research such as data mining and machine learning has emphasized the issues of class distributions in datasets. In typical predictive modeling, the inability to effectively address a class imbalance in a real-life dataset is an important shortcoming of the existing machine learning algorithms. Most algorithms assume a balanced class in their design, resulting in poor performance in predicting the minority target class. Ironically, the minority target class is usually the focus in predicting processes. The misclassification of the minority target class has resulted in serious consequences in detecting chronic diseases and detecting fraud and intrusion where positive cases are erroneously predicted as not positive. This paper presents a new attribute selection technique called variance ranking for handling imbalance class problems in a dataset. The results obtained were compared to two well-known attribute selection techniques: the Pearson correlation and information gain technique. This paper uses a novel similarity measurement technique ranked order similarity-ROS to evaluate the variance ranking attribute selection compared to the Pearson correlations and information gain. Further validation was carried out using three binary classifications: logistic regression, support vector machine, and decision tree. The proposed variance ranking and ranked order similarity techniques showed better results than the benchmarks. The ROS technique provided an excellent means of grading and measuring the similarities where other similarity measurement techniques were inadequate or not applicable.

Urban forest plays a crucial role in coping with global change and rapid urbanization via providing various ecosystem services. In this context, cities in Guangdong-Hongkong-Macao Greater Bay Area of China (GBA) established increasing investment in urban forest construction since 2000. However, data on the current condition of the planted forests remains scarce, which hampers a comprehensive understanding of how to improve the quality of urban forest planning and construction in the future. Therefore, we selected a core city Guangzhou, an inland city Foshan and a coastal city Zhuhai in GBA as our research objects and classified them into core urban area (CUA), semi urban area (SUA), and urban fringe area (UFA). We launched a field measurement for the tree variables regarding forest structure. Besides, ecological indexes were introduced to evaluate the condition and potential of the forests and florescence data together with tridimensional green quantity data provided different perspectives on landscape diversity and stability. The results show that Guangzhou has a well-planned forest pattern with advantages in forest structures and visible landscape values in different areas. To introduce different tree species in CUA and SUA and to plant more trees of those rare tree species in UFA can effectively increase Guangzhou's forest diversity and evenness. For Foshan, it is advisable to implement a balanced forest-investment policy by making specific recommendations for different areas, such as planting various tree species in CUA and UFA to increase forest diversity. For Zhuhai, it is distinguished for the attractiveness for tourists by the advantages in landscape diversity and stability, but it is still necessary to plant different trees with appropriate classes to enhance the urban forest stability. Overall, it is crucial for city managers to obtain precise knowledge of urban forest conditions from both the ecological and landscape perspectives, which may contribute enormously to the improvement of the quality of urban forest planning and construction in adapting to various environmental challenges.

Knowing the regeneration and structure of a species in different phytophysiognomies is extremely important to understand its pattern of occurrence. In spite of its importance, this type of study is still scarce in the Cerrado biome. In this perspective, this work aimed to compare the structure of the arboreal and regenerating strata of Annona coriacea in the cerrado sensu stricto and in the cerradão (savanna woodland). For that purpose, 13 plots of 400m² were installed in each physiognomy. All individuals of A. coriacea were sampled and had their diameter and height measured. The individuals were divided into regenerating and arboreal and were distributed in classes of height and diameter. Furthermore, the absolute density and the basal area of each extract were calculated in each physiognomy. The parameters were compared using the T-test. In total, 130 individuals of A. Coriacea were sampled, with 42 regenerating individuals in the cerrado sensu stricto and 49 in the cerradão. As for the arboreal individuals, 33 were sampled in the cerrado sensu stricto and 6 in the cerradão. All evaluated parameters revealed to be significantly different for the physiognomies. Based on these results, it is possible to affirm that A. coriacea did not present a distribution pattern in the two physiognomies, although it presents a better establishment success in the cerrado sensu stricto. In the cerradão, although the seeds can germinate, the seedlings fail to develop and reach the adult age. [ABSTRACT FROM AUTHOR]

Learning convolutional networks on graphs have been a popular topic for machine learning on graph-structured data and achieved state-of-the-art results on various practical tasks. However, most existing works ignore the impact of per-class distribution, therefore their performance may be limited due to the diversity of various categories. In this paper, we propose a novel class-aware progressive self-training (CPS) algorithm for training graph convolutional networks (GCNs). Compared to other self-training algorithms for GCNs' learning, the proposed CPS algorithm leverages the class distribution to update the original graph structure in each self-training loop, including: (a) find these high-confident unlabeled nodes in the graph for each category to add pseudo labels, in order to enlarge the current set of labeled nodes; (b) delete these noisy edges between different classes for graph sparsification. Then, the optimized graph is used for next self-training loops in hopes of enhancing the classification performance. We evaluate the proposed CPS on several datasets commonly used for GCNs' learning, and the experimental results show that the proposed CPS algorithm outperforms other baselines. [ABSTRACT FROM AUTHOR]

Highlights • Resolution enhancement considering the distortions due to the unstable UAV platforms. • Analysis of CNN architectures for modeling spatial distribution across spatial scales. • Investigation of the optimal network parameter settings for the proposed framework. • Illustration of the advantage of modeling spectral correlation for reconstruction. • Analysis of perceptual and image space loss functions for resolution enhancement. Abstract Spatial resolution enhancement is a pre-requisite for integrating unmanned aerial vehicle (UAV) datasets with the data from other sources. However, the mobility of UAV platforms, along with radiometric and atmospheric distortions, makes the task difficult. In this paper, various convolutional neural network (CNN) architectures are explored for resolving the issues related to sub-pixel classification and super-resolution of drone-derived datasets. The main contributions of this work are: 1) network-inversion based architectures for super-resolution and sub-pixel mapping of drone-derived images taking into account their spectral-spatial characteristics and the distortions prevalent in them 2) a feature-guided transformation for regularizing the inversion problem 3) loss functions for improving the spectral fidelity and inter-label compatibility of coarser to finer-scale mapping 4) use of multi-size kernel units for avoiding over-fitting. The proposed approach is the first of its kind in using neural network inversion for super-resolution and sub-pixel mapping. Experiments indicate that the proposed super-resolution approach gives better results in comparison with the sparse-code based approaches which generally result in corrupted dictionaries and sparse codes for multispectral aerial images. Also, the proposed use of neural network inversion, for projecting spatial affinities to sub-pixel maps, facilitates the consideration of coarser-scale texture and color information in modeling the finer-scale spatial-correlation. The simultaneous consideration of spectral bands, as proposed in this study, gives better super-resolution results when compared to the individual band enhancements. The proposed use of different data-augmentation strategies, for emulating the distortions, improves the generalization capability of the framework. Sensitivity of the proposed super-resolution and sub-pixel mapping frameworks with regard to the network parameters is thoroughly analyzed. The experiments over various standard datasets as well as those collected from known locations indicate that the proposed frameworks perform better when compared to the prominent published approaches. Graphical abstract Image, graphical abstract [ABSTRACT FROM AUTHOR]

In this paper we investigate the role of sample size and class distribution in credit risk assessments, focusing on real life imbalanced data sets. Choosing the optimal sample is of utmost importance for the quality of predictive models and has become an increasingly important topic with the recent advances in automating lending decision processes and the ever growing richness in data collected by financial institutions. To address the observed research gap, a large-scale experimental evaluation of real-life data sets of different characteristics was performed, using several classification algorithms and performance measures. Results indicate that various factors play a role in determining the optimal class distribution, namely the performance measure, classification algorithm and data set characteristics. The study also provides valuable insight on how to design the training sample to maximize prediction performance and the suitability of using different classification algorithms by assessing their sensitivity to class imbalance and sample size. [ABSTRACT FROM AUTHOR]

Highlights • Propose a new metric, LRID, to measure multi-class imbalance extent. • Discuss the effect of the number of minority classes on multi-class imbalance extent. • Demonstrate the effectiveness of LRID based on twenty UCI datasets for imbalanced learning. Abstract In this paper, we introduce a new likelihood ratio imbalance degree (LRID) to measure the class-imbalance extent of multi-class data. Imbalance ratio (IR) is usually used to measure class-imbalance extent in imbalanced learning problems. However, IR cannot capture the detailed information in the class distribution of multi-class data, because it only utilises the information of the largest majority class and the smallest minority class. Imbalance degree (ID) has been proposed to solve the problem of IR for multi-class data. However, we note that improper use of distance metric in ID can have harmful effect on the results. In addition, ID assumes that data with more minority classes are more imbalanced than data with less minority classes, which is not always true in practice. Thus ID cannot provide reliable measurement when the assumption is violated. In this paper, we propose a new metric based on the likelihood-ratio test, LRID, to provide a more reliable measurement of class-imbalance extent for multi-class data. Experiments on both simulated and real data show that LRID is competitive with IR and ID, and can reduce the negative correlation with F1 scores by up to 0.55. [ABSTRACT FROM AUTHOR]