Data set quality in Machine Learning: Consistency measure based on Group Decision Making

Metagenomics as a Transformative Tool for Antibiotic Resistance Surveillance: Highlighting the Impact of Mobile Genetic Elements with a Focus on the Complex Role of Phages

Extensive use of antibiotics in human healthcare as well as in agricultural and environmental settings has led to the emergence and spread of antibiotic-resistant bacteria, rendering many infections increasingly difficult to treat. Coupled with the limited development of new antibiotics, the rise of antimicrobial resistance (AMR) has caused a major health crisis worldwide, which calls for immediate action. Strengthening AMR surveillance systems is, therefore, crucial to global and national efforts in combating this escalating threat. This review explores the potential of metagenomics, a sequenced-based approach to analyze entire microbial communities without the need for cultivation, as a transformative and rapid tool for improving AMR surveillance strategies as compared to traditional cultivation-based methods. We emphasize the importance of monitoring mobile genetic elements (MGEs), such as integrons, transposons, plasmids, and bacteriophages (phages), in relation to their critical role in facilitating the dissemination of genetic resistance determinants via horizontal gene transfer (HGT) across diverse environments and clinical settings. In this context, the strengths and limitations of current bioinformatic tools designed to detect AMR-associated MGEs in metagenomic datasets, including the emerging potential of predictive machine learning models, are evaluated. Moreover, the controversial role of phages in AMR transmission is discussed alongside the potential of phage therapy as a promising alternative to conventional antibiotic treatment.

BDMANGO: An image dataset for identifying the variety of mango based on the mango leaves

In the field of agriculture, particularly within the context of machine learning applications, quality datasets are essential for advancing research and development. To address the challenges of identifying different mango leaf types and recognizing the diverse and unique characteristics of mango varieties in Bangladesh, a comprehensive and publicly accessible dataset titled "BDMANGO" has been created. This dataset includes images essential for research, featuring six mango varieties: Amrapali, Banana, Chaunsa, Fazli, Haribhanga, and Himsagar, which were collected from different locations. The images were captured using the rear cameras of a Google Pixel 6a and an iPhone XR and were stored in 640 × 480 pixels resolution. Both sides of each mango leaf were photographed against white background to accurately reflect real-world scenarios in mango cultivation fields. The white background was specifically chosen to remove noise in image sample, allowing for accurate feature extraction by machine learning algorithms. This will ensure the trained model's efficacy in identifying a specific mango leaf while implemented alongside any segmentation algorithm. Additionally, image augmentation techniques such as rotation, horizontal flip, vertical flip, width shift, height shift, shear range, and zooming were applied to expand the dataset from 837 original images to a total of 6696 images (837 original image and 5859 augmented images). This expansion significantly enhances the dataset's utility for training, testing, and validating machine learning models designed for classifying mango leaf varieties, thereby supporting research efforts in this domain.

The Venus score for the assessment of the quality and trustworthiness of biomedical datasets

Biomedical datasets are the mainstays of computational biology and health informatics projects, and can be found on multiple data platforms online or obtained from wet-lab biologists and physicians. The quality and the trustworthiness of these datasets, however, can sometimes be poor, producing bad results in turn, which can harm patients and data subjects. To address this problem, policy-makers, researchers, and consortia have proposed diverse regulations, guidelines, and scores to assess the quality and increase the reliability of datasets. Although generally useful, however, they are often incomplete and impractical. The guidelines of Datasheets for Datasets, in particular, are too numerous; the requirements of the Kaggle Dataset Usability Score focus on non-scientific requisites (for example, including a cover image); and the European Union Artificial Intelligence Act (EU AI Act) sets forth sparse and general data governance requirements, which we tailored to datasets for biomedical AI. Against this backdrop, we introduce our new Venus score to assess the data quality and trustworthiness of biomedical datasets. Our score ranges from 0 to 10 and consists of ten questions that anyone developing a bioinformatics, medical informatics, or cheminformatics dataset should answer before the release. In this study, we first describe the EU AI Act, Datasheets for Datasets, and the Kaggle Dataset Usability Score, presenting their requirements and their drawbacks. To do so, we reverse-engineer the weights of the influential Kaggle Score for the first time and report them in this study. We distill the most important data governance requirements into ten questions tailored to the biomedical domain, comprising the Venus score. We apply the Venus score to twelve datasets from multiple subdomains, including electronic health records, medical imaging, microarray and bulk RNA-seq gene expression, cheminformatics, physiologic electrogram signals, and medical text. Analyzing the results, we surface fine-grained strengths and weaknesses of popular datasets, as well as aggregate trends. Most notably, we find a widespread tendency to gloss over sources of data inaccuracy and noise, which may hinder the reliable exploitation of data and, consequently, research results. Overall, our results confirm the applicability and utility of the Venus score to assess the trustworthiness of biomedical data.

Toward High-Quality Real-World Laboratory Data in the Era of Healthcare Big Data

With Industry 4.0, big data and artificial intelligence have become paramount in the field of medicine. Electronic health records, the primary source of medical data, are not collected for research purposes but represent real-world data; therefore, they have various constraints. Although structured, laboratory data often contain unstandardized terminology or missing information. The major challenge lies in the lack of standardization of test results in terms of metrology, which complicates comparisons across laboratories. In this review, we delve into the essential components necessary for integrating real-world laboratory data into high-quality big data, including the standardization of terminology, data formats, equations, and the harmonization and standardization of results. Moreover, we address the transference and adjustment of laboratory results, along with the certification for quality of laboratory data. By discussing these critical aspects, we seek to shed light on the challenges and opportunities inherent to utilizing real-world laboratory data within the framework of healthcare big data and artificial intelligence.

Optimization of SVR and CatBoost models using metaheuristic algorithms to assess landslide susceptibility

In this study, a landslide susceptibility assessment is performed by combining two machine learning regression algorithms (MLRA), such as support vector regression (SVR) and categorical boosting (CatBoost), with two population-based optimization algorithms, such as grey wolf optimizer (GWO) and particle swarm optimization (PSO), to evaluate the potential of a relatively new algorithm and the impact that optimization algorithms can have on the performance of regression models. The Kerala state in India has been chosen as the test site due to the large number of recorded incidents in the recent past. The study started with 18 potential predisposing factors, which were reduced to 14 after a multi-approach feature selection technique. Six susceptibility models were implemented and compared using the machine learning algorithms alone and combining each of them with the two optimization algorithms: SVR, CatBoost, SVR-PSO, CatBoost-PSO, SVR-GWO, and CatBoost-GWO. The resulting maps were validated with an independent dataset. The performance rankings, based on the area under the receiver operating characteristic curve (AUC) metric, are as follows: CatBoost-GWO (AUC = 0.910) had the highest performance, followed by CatBoost-PSO (AUC = 0.909), CatBoost (AUC = 0.899), SVR-GWO (AUC = 0.868), SVR-PSO (AUC = 0.858), and SVR (AUC = 0.840). Other validation statistics corroborated these outcomes, and the Friedman and Wilcoxon-signed rank tests verified the statistical significance of the models. Our case study showed that CatBoost outperformed SVR both in case the models were optimized or not; the introduction of optimization algorithms significantly improves the results of machine learning models, with GWO being slightly more effective than PSO. However, optimization cannot drastically alter the results of the model, highlighting the importance of setting up of a rigorous susceptibility model since the early steps of any research.

Learning-based real-time imaging through dynamic scattering media

Imaging through dynamic scattering media is one of the most challenging yet fascinating problems in optics, with applications spanning from biological detection to remote sensing. In this study, we propose a comprehensive learning-based technique that facilitates real-time, non-invasive, incoherent imaging of real-world objects through dense and dynamic scattering media. We conduct extensive experiments, demonstrating the capability of our technique to see through turbid water and natural fog. The experimental results indicate that the proposed technique surpasses existing approaches in numerous aspects and holds significant potential for imaging applications across a broad spectrum of disciplines.

Binary dataset for machine learning applications to tropical cyclone formation prediction

Applications of machine learning (ML) in atmospheric science have been rapidly growing. To facilitate the development of ML models for tropical cyclone (TC) research, this binary dataset contains a specific customization of the National Center for Environmental Prediction (NCEP)/final analysis (FNL) data, in which key environmental conditions relevant to TC formation are extracted for a range of lead times (0-72 hours) during 1999-2023. The dataset is designed as multi-channel images centered on TC formation locations, with a positive and negative directory structure that can be readily read from any ML applications or common data interface. With its standard structure, this dataset provides users with a unique opportunity to conduct ML application research on TC formation as well as related predictability at different forecast lead times.

Water Effective Diffusion Coefficient in Dairy Powder Calculated by Digital Image Processing and through Machine Learning Algorithms of CLSM Micrographs

Rehydration of dairy powders is a complex and essential process. A relatively new quantitative mechanism for monitoring powders' rehydration process uses the effective diffusion coefficient. This research focused on modifying a previously used labor-intensive method that will be able to automatically measure the real-time water diffusion coefficient in dairy powders based on confocal microscopy techniques. Furthermore, morphological characteristics and local hydration of individual particles were identified using an imaging analysis procedure written in Matlab©-R2023b and image analysis through machine learning algorithms written in Python™-3.11. The first model includes segmentation into binary images and labeling particles during water diffusion. The second model includes the expansion of data set selection, neural network training and particle markup. For both models, the effective diffusion follows Fick's second law for spherical geometry. The effective diffusion coefficient on each particle was computed from the dye intensity during the rehydration process. The results showed that effective diffusion coefficients for water increased linearly with increasing powder particle size and are in agreement with previously used methods. In summary, the models provide two independent machine measurements of effective diffusion coefficient based on the same set of micrographs and may be useful in a wide variety of high-protein powders.

A review on advancements in lithological mapping utilizing machine learning algorithms and remote sensing data

Lithological mapping is a fundamental undertaking in geological research, mineral resource exploration, and environmental management. However, conventional methods for lithological mapping are often laborious and challenging, particularly in remote or inaccessible areas. Fortunately, a transformative solution has emerged through the integration of remote sensing and machine learning algorithms, providing an efficient and accurate means of deciphering the geological features of the Earth's crust. Remote sensing offers vast and comprehensive data across extensive geographical regions, while machine learning algorithms excel at recognizing intricate patterns and features in the data, enabling the classification of different lithological units. Compared to traditional methods, this approach is faster, more efficient, and remarkably accurate. The combination of remote sensing and machine learning presents numerous advantages, including the ability to amalgamate multiple data sources, provide up-to-date information on rapidly changing regions, and manage vast volumes of data. This review article delves into the invaluable contributions of remote sensing and machine learning algorithms to lithological mapping. It extensively explores diverse remote sensing datasets, such as Landsat, Sentinel-2, ASTER, and Hyperion data, which can be effectively harnessed for this purpose. Additionally, the study investigates a range of machine learning algorithms, including SVM, RF, and ANN, specifically tailored for lithological mapping. By scrutinizing practical use cases, the review underscores the strengths, limitations, and potential future developments of remote sensing and machine learning algorithms in the context of lithological mapping. Practical use cases have demonstrated the immense potential of machine learning algorithms, with the SVM classifier emerging as a reliable and accurate option for lithological mapping. Moreover, the choice of the most appropriate data source depends on the specific objectives of the application. Overall, the transformative potential of remote sensing and machine learning in lithological mapping cannot be overstated. This integrated approach not only enhances our understanding of geological features but also enables diverse applications in geological research and environmental management. With the promise of a more informed and sustainable future, the utilization of remote sensing and machine learning in lithological mapping represents a pivotal advancement in the field of geological sciences.

Toward reliable machine learning with Congruity: a quality measure based on formal concept analysis

The spreading of machine learning (ML) and deep learning (DL) methods in different and critical application domains, like medicine and healthcare, introduces many opportunities but raises risks and opens ethical issues, mainly attaining to the lack of transparency. This contribution deals with the lack of transparency of ML and DL models focusing on the lack of trust in predictions and decisions generated. In this sense, this paper establishes a measure, namely Congruity, to provide information about the reliability of ML/DL model results. Congruity is defined by the lattice extracted through the formal concept analysis built on the training data. It measures how much the incoming data items are close to the ones used at the training stage of the ML and DL models. The general idea is that the reliability of trained model results is highly correlated with the similarity of input data and the training set. The objective of the paper is to demonstrate the correlation between the Congruity and the well-known Accuracy of the whole ML/DL model. Experimental results reveal that the value of correlation between Congruity and Accuracy of ML model is greater than 80% by varying ML models.