High Error Rate Research Articles

NanoMGT: Marker gene typing of low complexity mono-species metagenomic samples using noisy long reads.

Rapid advancements in sequencing technologies have led to significant progress in microbial genomics, yet challenges persist in accurately identifying microbial strain diversity in metagenomic samples, especially when working with noisy long-read data from platforms like Oxford Nanopore Technologies (ONT). In this article, we introduce NanoMGT, a tool designed to enhance marker gene typing in low-complexity mono-species samples, leveraging the unique properties of long reads. NanoMGT excels in its ability to accurately identify mutations amidst high error rates, ensuring the reliable detection of multiple strain-specific marker genes. Our tool implements a novel scoring system that rewards mutations co-occurring across different reads and penalizes densely grouped, likely erroneous variants, thereby achieving a good balance between sensitivity and precision. A comparative evaluation of NanoMGT, using a simulated multi-strain sample of seven bacterial species, demonstrated superior performance relative to existing tools and the advantages of using a threshold-based filtering approach to calling minority variants in ONT's sequencing data. NanoMGT's potential as a post-binning tool in metagenomic pipelines is particularly notable, enabling researchers to more accurately determine specific alleles and understand strain diversity in microbial communities. Our findings have significant implications for clinical diagnostics, environmental microbiology, and the broader field of genomics. The findings offer a reliable and efficient approach to marker gene typing in complex metagenomic samples.

Overcoming the High Error Rate of Composite DNA Letters-Based Digital Storage through Soft-Decision Decoding.

Composite DNA letters, by merging all four DNA nucleotides in specified ratios, offer a pathway to substantially increase the logical density of DNA digital storage (DDS) systems. However, these letters are susceptible to nucleotide errors and sampling bias, leading to a high letter error rate, which complicates precise data retrieval and augments reading expenses. To address this, Derrick-cp is introduced as an innovative soft-decision decoding algorithm tailored for DDS utilizing composite letters. Derrick-cp capitalizes on the distinctive error sensitivities among letters to accurately predict and rectify letter errors, thus enhancing the error-correcting performance of Reed-Solomon codes beyond traditional hard-decision decoding limits. Through comparative analyses in the existing dataset and simulated experiments, Derrick-cp's superiority is validated, notably halving the sequencing depth requirement and slashing costs by up to 22% against conventional hard-decision strategies. This advancement signals Derrick-cp's significant role in elevating both the precision and cost-efficiency of composite letter-based DDS.

Anti-noise performance analysis in amplitude-modulated collinear holographic data storage using deep learning

In an amplitude-modulated collinear holographic data storage system, optical system aberration and experimental noise due to the recording medium often result in a high bit error rate (BER) and low signal-to-noise ratio (SNR) in directly read detector data. This study proposes an anti-noise performance analysis using deep learning. End-to-end convolutional neural networks were employed to analyze noise resistance in encoded data pages captured by the detector. Experimental results demonstrate that these networks effectively correct system imaging aberrations, detector light intensity response, holographic storage medium response non-uniformity, and defocusing noise from the recording objective lens. Consequently, the BER of reconstructed encoded data pages can be reduced to 1/10 of that from direct detection, while the SNR can be increased more than fivefold, enhancing the accuracy and reliability of data reading in amplitude holographic data storage systems.

Tracking the impact of heavy metals on human health and ecological environments in complex coastal aquifers using improved machine learning optimization.

The rising heavy metal (HM) pollution in coastal aquifers in rapidly urbanizing areas such as Dammam leads to significant risks to public health and environmental sustainability, challenging compliance with Environmental Protection Agency (EPA) guidelines, World Health Organization (WHO) standards, and Sustainable Development Goals (SDGs) related to clean water and life on land. This study developed the predictive-based monitoring of HM concentrations, including cadmium (Cd), chromium (Cr), and mercury (Hg) in the coastal aquifers of Dammam, influenced by industrial, agricultural, and urban activities. For this purpose, dynamic system identification and machine learning (ML) models integrated with three ensemble techniques, namely, simple averaging (SAE), weighted averaging (WAE), and neuro-ensemble (N-ESB), were employed to enhance the accuracy, reliability, and efficiency of environmental monitoring systems. The experimental data were calibrated and validated in addition to k-fold cross-validation to ensure the predictive skills of the models. The methodology integrates extensive data collection across varied land uses in Dammam and accurate model calibration and validation phases to develop highly accurate predictive models. The findings proved that the N-ESB and Hammerstein-Wiener (HW) models surpassed other models in predicting the concentrations of all HM. For Cd, the N-ESB model achieved a root mean square error (RMSE = 0.0010mg/kg). Similarly, Cr demonstrated superior performance (RMSE = 0.0179mg/kg). Further numerical results indicated that the HW algorithm proved the most effective for Hg, with RMSE = 0.0000mg/kg. The quantitative comparison suggested that the N-ESB model's consistently high performance and low error rates make it an optimal choice for real-time, precise monitoring and management of HM pollution in coastal aquifers. The outcomes of this research highlighted the importance of integrating advanced predictive modeling techniques in environmental science, providing significant and practical implications for policymaking and ecological management.

Genotyping SNPs and Indels: A method to improve the scope and sensitivity of High-Resolution melt (HRM) analysis based applications

High-resolution melt (HRM) analysis is a closed-tube technique for detecting single nucleotide polymorphisms (SNPs). However, it has limited use in high-resolution melting devices, even those with high thermal accuracy (HTA). In addition to the cost of switching to these specialized devices, the presence of nearest neighbour neutral changes (class III, IV SNPs and small indels) made HRM-based assays a challenging task due to reduced sensitivity. This study aimed to design a common modified competitive amplification of differently melting amplicons (CADMA)-based assay to address these challenges by generating allele-specific qPCR products that are detectable on most qPCR platforms.For this study, SNPs were selected from all four classes of SNPs (class I: C/T or G/A mutation; class II: C/A or G/T mutation; class III: G/C mutation; class IV: A/T mutation). A single base pair and 19 bp indels were also chosen to simulate how CADMA primers could be designed for indels of varying lengths. The melting temperatures (Tm) were determined using IDT oligoAnalyzer. qPCR and melt data acquisition were performed on the CFX96 qPCR platform, and the melt curve data were analyzed using Precision Melt software (Bio-Rad, USA). The clusters for different genotypes were successfully identified with the aid of the control samples, and Tm predictions were carried out using the uMelt batch and Tm online tools for comparison.Using HRM-qPCR assays based on the modified CADMA method, genotyping of various SNPs was successfully carried out. For some SNPs, similarly shaped melt curves were observed for homozygotes and heterozygotes, making shape-based genotype prediction difficult. The Tm values calculated via the Blake and Delcourts (1998) method were the closest to the experimental Tm values after adjusting for the salt concentration.Since HRM assays usually depend on the ΔTm caused by mutations, they are prone to a high error rate due to nearest neighbour neutral changes. The technique developed in this study significantly reduces the failure rates in HRM-based genotyping and could be applied to any SNP or indel in any platform. It is crucial to have a deep understanding of the melt instrument, its accuracy and the nature of the target (SNP class or indel length and GC content of the flanking region). Furthermore, the availability of controls is essential for a high success rate.

Comparative Evaluation of Colistin-Susceptibility Testing in Carbapenem-Resistant Klebsiella pneumoniae Using VITEK, Colistin Broth Disc Elution, and Colistin Broth Microdilution.

The study aimed to compare the results of colistin-susceptibility testing performed using the automated VITEK system, colistin broth microdilution (BMD), and colistin broth disk elution (CBDE) methods. This exploratory study was conducted in a tertiary care center in South India. Carbapenem-resistant Klebsiella pneumoniae (n = 49) isolates collected from a clinical microbiology laboratory over six months (March-September 2023) were used for the study. Among the 49 carbapenem-resistant Klebsiella pneumoniae isolates, 42 were found to be susceptible to carbapenem by all three methods. Seven isolates were found to be resistant to colistin using BMD and CBDE methods. Two isolates were incorrectly detected as colistin-susceptible, and one isolate was wrongly categorized as colistin-resistant using the automated VITEK system. CBDE is a reliable and cost-effective method that can be adopted in the routine microbiology laboratory for colistin-susceptibility testing, as it does not require any specialized equipment or techniques and is 100% consistent with the gold standard BMD method. Although the automated VITEK system is used in most routine microbiological laboratories for antibiotic-susceptibility testing, it cannot be reliably used for colistin-susceptibility testing due to its high error rates (very major error rate of 28.5%; major error rate of 2.4%).

Using AuNPs-DNA Walker with Fluorophores Detects the Hepatitis Virus Rapidly.

Viral hepatitis is a systemic infectious diseases caused by various hepatitis viruses, primarily leading to liver damage. It is widely prevalent worldwide, with hepatitis viruses categorized into five types: hepatitis A, B, C, D, and E, based on their etiology. Currently, the detection of hepatitis viruses relies on methods such as enzyme-linked immunosorbent assay (ELISA), immunoelectron microscopy to observe and identify viral particles, and in situ hybridization to detect viral DNA in tissues. However, these methods have limitations, including low sensitivity, high error rates in results, and potential false negative reactions due to occult serum infection conditions. To address these challenges, we have designed an AuNPs-DNA walker method that uses gold nanoparticles (AuNPs) and complementary DNA strands for detecting viral DNA fragments through a colorimetric assay and fluorescence detection. The DNA walker, attached to gold nanoparticles, comprises a long walking strand with a probe sequence bound and stem-loop structural strands featuring a modified fluorescent molecule at the 3' end, which contains the DNAzyme structural domain. Upon the addition of virus fragments, the target sequence binds to the probe chains. Subsequently, the long walking strand is released and continuously hybridizes with the stem-loop structural strand. The DNAzyme undergoes hydrolytical cleavage by Mg2+, breaking the stem-loop structural strand into linear single strands. As a result of these structural changes, the negative charge density in the solution decreases, weakening spatial repulsion and rapidly reducing the stability of the DNA walker. This leads to aggregation upon the addition of a high-salt solution, accompanied by a color change. Virus typing can be performed through fluorescence detection. The innovative method can detect DNA/RNA fragments with high specificity for the target sequence, reaching concentrations as low as 1 nM. Overall, our approach offers a more convenient and reliable method for the detection of hepatitis viruses.

Pediatric asthma inhaler technique: quality and content analysis of YouTube videos

Background Proper technique for using inhalers is crucial in treating pediatric asthma. YouTube offers a wide range of videos on pediatric inhaler technique, but there is a need to analyze the quality, reliability, and content of these resources. Aims This study aims to analyze the quality, reliability, and content of YouTube videos on pediatric asthma inhaler techniques. Methods The study has a descriptive, retrospective, and cross-sectional design. The research was conducted by searching YouTube using the “Pediatric Metered Dose Inhaler,” “Pediatric Accuhaler,” and “Pediatric Diskus.” The video’s popularity was measured using the Video Power Index. The quality and reliability of the videos were evaluated using the modified DISCERN and Global Quality Scale (GQS). Results This study analyzed 55 YouTube videos on the pediatric inhaler technique. 19 of the videos were related to the pMDI inhaler with a spacer for tidal breathing, 14 pMDI inhaler with a spacer for single breath, and 22 diskus device. Findings show that videos demonstrating the use of pMDI devices for single breath have more reliable modified DISCERN scores. However, videos related to tidal breathing are more popular than those showing the use of diskus devices and single breath. Based on the checklist for videos on diskus devices, the steps with the highest error rates are ‘Check dose counter’ at 72.7% and ‘Breathe out gently, away from the inhaler’ at 63.6%. A moderate correlation was observed between the modified DISCERN score and the GQS. Conclusions While YouTube videos on the pMDI single-breath technique may be useful for pediatric patients and caregivers, it is crucial for them to receive inhaler technique education from their healthcare provider. This study’s findings hold great significance for pediatric patients and caregivers, particularly those who rely on YouTube for health-related information.

Judging the algorithm

AbstractThis paper discusses an algorithmic tool introduced in the Basque Country (Spain) to assess the risk of intimate partner violence. The algorithm was introduced to address the lack of human experts by automatically calculating the level of violence based on psychometric features such as controlling or violent behaviour. Given that critical literature on risk assessment tools for domestic violence mainly focuses on English-speaking countries, this paper offers an algorithmic accountability analysis in a non-English speaking region. It investigates the algorithmic risks, harms, and limitations associated with the Basque tool. We propose a transdisciplinary approach from a critical statistical and legal perspective. This approach unveils issues and limitations that could lead to unexpected consequences for individuals suffering from partner violence. Moreover, our analysis suggests that the algorithmic tool has a high error rate on severe cases, i.e., cases where the aggressor could murder his partner—5 out of 10 high-risk cases are misclassified as low risk—and that there is a lack of appropriate legal guidelines for judges, the end users of this tool. The paper concludes that this risk assessment tool needs to be urgently evaluated by independent and transdisciplinary experts to better mitigate algorithmic harms in the context of intimate partner violence.

What's old is new: leveraging existing antimicrobial susceptibility test methods for rapid results in patients with bloodstream infections.

The use of rapid disk diffusion or modified automated antimicrobial susceptibility testing (AST) system approaches demonstrates excellent performance for gram-negative organisms directly from blood cultures. In a recent study, S. Khan, A. Das, A. Mishra, A. Vidyarthi, et al. (Microbiol Spectr 12:e03081-23, 2024, https://doi.org/10.1128/spectrum.03081-23) compared the performance of three direct-from-blood AST methods against standard of care disk diffusion and automated AST. The results demonstrated high categorical agreements and low error rates across three protocols. The study suggests that locally validated direct-from-blood AST protocols offer reliable and fast results, particularly for resource-limited settings. However, local context and workflows should be considered prior to implementing rapid AST protocols, and more research is needed on the performance of rapid AST protocols for gram-positive organisms.

An Error Protection Protocol for the Multicast Transmission of Data Samples in V2X Applications

There is a trend towards communication of larger data objects in wireless vehicle communication. In many cases, communication uses publish-subscribe protocols. Data rate requirements of such protocols are best addressed by wireless multicast protocols, but the existing protocols lack an error protection that is suitable for real-time and safety-critical applications. We present an application-aware protocol that supports the popular DDS (Data Distribution Service) middleware. By exploiting data object deadlines and slack for retransmissions and employing an adaptable, multicast-aware prioritization mechanism, the reliable exchange of large data objects is enabled. The protocol is sufficiently general to be used on top of different communication standards such as 802.11- and cellular-based V2X (Vehicle-to-Everything) technologies. The protocol was implemented in an OMNeT++ simulation model and evaluated against recent state-of-the-art alternatives using parameters and constraints taken from a motivational truck platooning example. Furthermore, the protocol was implemented using an open-source DDS implementation as the basis and tested on a physical wireless demonstrator setup. The evaluation shows that the presented multicast protocol substantially outperforms the alternatives keeping streaming applications operational even under high frame error rates.

Fingerphoto morphing attack generation using texture descriptors based landmarks

Smartphone-based biometric authentication has been widely used in various applications. Among several biometric characteristics, fingerphoto biometrics captured from smartphones are gaining popularity owing to their usability, scalability across different smartphones, and reliable verification. However, fingerphoto verification systems are vulnerable to both direct and indirect attacks. In this work, we propose a novel method to generate morphing attacks on fingerphoto biometrics captured using smartphones. We introduce three different image-level fingerphoto morphing attack generation algorithms that can generate high-quality fingerphoto morphing images with minimum distortions. Extensive experiments were conducted on two datasets captured using different smartphones under various environmental conditions. The results demonstrate that the proposed morphing algorithms are highly vulnerable to commercial off-the-shelf and block-directional fingerprint verification systems. To effectively detect morphing attacks on fingerphoto biometrics, we propose the use of fingerphoto morphing attack detection algorithms that utilize both handcrafted and deep features. However, our detection results showed a high error rate in accurately detecting these types of attacks.

Mitigating Temporal Fragility in the XY Surface Code

An important outstanding challenge that must be overcome in order to fully utilize the XY surface code for correcting biased Pauli noise is the phenomenon of fragile temporal boundaries that arises during the standard logical state-preparation and measurement protocols. To address this challenge we propose a new logical state-preparation protocol based on locally entangling qubits into small Greenberger-Horne-Zeilinger-like states prior to making the stabilizer measurements that place them in the XY-code state. We prove that in this new procedure O(n) high-rate errors along a single lattice boundary can cause a logical failure, leading to an almost quadratic reduction in the number of fault configurations compared to the standard state-preparation approach. Moreover, the code becomes equivalent to a repetition code for high-rate errors, guaranteeing a 50% code-capacity threshold during state preparation for infinitely biased noise. With a simple matching decoder we confirm that our preparation protocol outperforms the standard protocol in terms of both threshold and logical error rate in the fault-tolerant regime where measurements are unreliable and at experimentally realistic biases. We also discuss how our state-preparation protocol can be inverted for similar fragile-boundary-mitigated logical-state measurement. Published by the American Physical Society 2024

Coupling Coherent Point Drift And Affine Least-Squares Transformation To Build Trajectories In TR-PTV

The robustness and accuracy of state-of-the-art time-resolved Particle Tracking Velocimetry algorithms is strongly supported by the use of the temporal coherence of particle trajectories. However, to exploit this coherence, the particle needs to be tracked for a sufficiently long period. This condition cannot be achieved if the time of flight of the particles across the measurement volume is not long enough, for example because the limited illumination energy requires the thickness of the volume to be thin. The authors propose to address this difficulty by providing an algorithm based on the spatial coherence of the trajectories rather than the temporal one. This algorithm results from the coupling of two methods; the Coherent Point Drift (CPD), which has proven to be accurate in pairing particles from two instants, but with an insufficient recall for time-resolved application, and the Affine Least Square Transformation (ALST), which allows transporting particles that were missed by the CPD. This algorithm is integrated into a global process that includes an Iterative Particle Reconstruction (IPR) algorithm to extract a list of particles from the images. First tests performed on a synthetic test case consisting of images of particles advected by a turbulent flow highlighted the potential of the method for tracking short tracks. With six frames, the recall is found to be 99.6% with the present method against 98.3% with the DaVis 10.2 implementation of ShakeThe-Box (STB). However, the error is much larger with a false positive rate of 5.1% with our method but only 0.2% with (STB). The cause of this high error rate is analysed and it is speculated that it is partly explained by the tendency of the IPR to detect the same real particle multiple times, which is a concern that has to be addressed in future work.

Advancements in ultrasonography and MRI for enhanced diagnosis of Parkinson's disease: An updated review

Parkinson's disease (PD) is a progressive neurological illness characterized by symptoms such as bradykinesia (slowness of movement), tremor, rigidity, and postural instability. Diagnosing parkinsonian syndromes (PS) can be challenging, even for specialized neurology centers, due to the high rate of diagnostic errors. This review article highlights the challenges in diagnosing PS and the importance of accurately differentiating it from PD. It discusses the utility of various magnetic resonance imaging (MRI) techniques and transcranial ultrasound (TUS) in improving diagnostic accuracy. The focus is on both conventional and cutting-edge MRI procedures, including T2*/T2-weighted MRI, susceptibility-weighted imaging (SWI), diffusion-weighted imaging (DWI), diffusion tensor imaging (DTI), magnetization transfer imaging (MTI), magnetic resonance spectroscopy (MRS), and ultra-high-resolution MRI. These advanced techniques have shown promising potential in aiding diagnosis and identifying early signs of PD. The review emphasizes the clinical relevance of these imaging modalities in discriminating PD from other parkinsonian disorders and detecting early-stage disease before the onset of clinical symptoms.

Automated Processing Method for Chinese NOTAMs Based on Knowledge Graph

Notice to airmen (NOTAMs) constitutes a vital element in civil aviation operational intelligence. Historically, the processing of these notices has been manual. However, with the significant increase in the number of NOTAMs, issues including low efficiency, time-consuming processes, and high error rates associated with manual processing have become apparent. To address these challenges, we propose an enhanced approach utilizing the Bi-GRU-CRF-Attention model, based on a dataset of 105,797 NOTAMs collected from the Intelligence Center between September 2020 and April 2023. In this methodology, we employ preprocessing techniques to train the model using processed NOTAMs. Subsequently, the trained model is utilized for named entity recognition, identifying entities within the notices, such as status, facilities, and reasons and segmenting sentences into words. Following this, an advanced BERT-DPCNN method is employed to classify the identified entities, yielding triplets comprising NOTAM entities, their categories, and corresponding processing methods. By integrating rule-based approaches, we configure a NOTAM knowledge graph using neo4j. This process establishes an automated NOTAM processing system. This system can autonomously determine the category of a NOTAM upon reception and utilize the Cypher language to query for the appropriate processing method.

Challenges in estimating species' age from phylogenetic trees

AbstractAimSpecies age, the elapsed time since origination, can give insight into how species longevity might influence eco‐evolutionary dynamics, which has been hypothesized to influence extinction risk. Traditionally, species' ages have been estimated from fossil records. However, numerous studies have recently used the branch lengths of time‐calibrated phylogenies as estimates of the ages of extant species. This approach poses problems because phylogenetic trees only contain direct information about species identity at the tips and not along the branches. Here, we show that incomplete taxon sampling, extinction and different assumptions about speciation modes can significantly alter the relationship between true species age and phylogenetic branch lengths, leading to high error rates. We found that these biases can lead to erroneous interpretations of eco‐evolutionary patterns derived from comparing phylogenetic age and other traits, such as extinction risk.InnovationFor bifurcating speciation, the default assumption in most analyses of species age, we propose a probabilistic approach based on the properties of a birth–death process to improve the estimation of species ages. Our approach can reduce the error by one order of magnitude under cases of high extinction and a high percentage of unsampled extant species.Main conclusionOur results call for caution in interpreting the relationship between phylogenetic ages and eco‐evolutionary traits, as this can lead to biased and erroneous conclusions. We show that, under the assumption of bifurcating speciation, we can obtain unbiased approximations of species age by combining information from branch lengths with the expectations of a birth–death process.

대조언어학적 관점을 통한 베트남어권 한국어 학습자를 위한 비음화 오류양상 고찰과 시사점

Objectives The purpose of this study is to obtain educational implications by examining the patterns of nasalization errors (correct and incorrect answers) that appear in each environment among Vietnamese Korean language learners. Methods In order to proceed with this discussion, we first reviewed previous research and theoretical background, and then established standards to analyze the nasalization error rate by environment. After that, from January 30, 2023 to January 30, 2024, in Korea, 19 to 30-year-olds After a basic survey, 15 Vietnamese Korean language learners (5 men and 10 women each) met 1:1 via Zoom and were asked to utter pre-prepared words related to nasal pronunciation. Then, by determining whether the answer was incorrect or correct, more detailed error types and their characteristics according to the overall results and characteristics of each environment were extracted. Results As a result of the analysis, through a review of previous research related to contrastive linguistics, it was possible to understand the fundamental cause of nasalization through differences in syllable structure between the two languages, and through a basic survey, despite the high demand for nasalization education, there was no awareness due to lack of educational experience. I could see that it was lacking. In addition, the results of error patterns by environment showed that the highest rate was in the case of ‘nasalization after adding ‘ㄴ’ or ‘nasalization after consonant group simplification.’ It was found that this was caused by a lack of awareness of nasalization by environment among the nasalization rules of Korean, the standard language. In addition, the higher error rate at ‘word boundaries’ rather than ‘word interior’ showed that nasalization education should be done ‘by phrase’. Lastly, as a result of examining error patterns by region, grade, and gender in an environment that previously showed a high error rate, all grades showed similar difficulties in pronunciation of bar consonants, but by region, it was more common in the south than in the north and central regions, and by gender, more than in women. A high error rate was observed in men. Conclusions In this way, this study identified the actual status of nasalization errors (total error rate by environment and error rate by word interior/boundary) of Vietnamese learners in more diverse environments and, unlike existing previous studies, provided basic data and implications for preparing educational measures. It was a meaningful study in that it was able to obtain results.

Optimization-based convolutional neural model for the classification of white blood cells

White blood cells (WBCs) are one of the most significant parts of the human immune system, and they play a crucial role in diagnosing the characteristics of pathologists and blood-related diseases. The characteristics of WBCs are well-defined based on the morphological behavior of their nuclei, and the number and types of WBCs can often determine the presence of diseases or illnesses. Generally, there are different types of WBCs, and the accurate classification of WBCs helps in proper diagnosis and treatment. Although various classification models were developed in the past, they face issues like less classification accuracy, high error rate, and large execution. Hence, a novel classification strategy named the African Buffalo-based Convolutional Neural Model (ABCNM) is proposed to classify the types of WBCs accurately. The proposed strategy commences with collecting WBC sample databases, which are preprocessed and trained into the system for classification. The preprocessing phase removes the noises and training flaws, which helps improve the dataset's quality and consistency. Further, feature extraction is performed to segment the WBCs, and African Buffalo fitness is updated in the classification layer for the correct classification of WBCs. The proposed framework is modeled in Python, and the experimental analysis depicts that it achieved 99.12% accuracy, 98.16% precision, 99% sensitivity, 99.04% specificity, and 99.02% f-measure. Furthermore, a comparative assessment with the existing techniques validated that the proposed strategy obtained better performances than the conventional models.

Development of advanced machine learning for prognostic analysis of drying parameters for banana slices using indirect solar dryer

In this study, eXtreme Gradient Boosting (XGBoost) and Light Gradient Boosting (LightGBM) algorithms were used to model-predict the drying characteristics of banana slices with an indirect solar drier. The relationships between independent variables (temperature, moisture, product type, water flow rate, and mass of product) and dependent variables (energy consumption and size reduction) were established. For energy consumption, XGBoost demonstrates superior performance with an R2 of 0.9957 during training and 0.9971 during testing, alongside minimal MSE of 0.0034 during training and 0.0008 during the testing phase indicating high predictive accuracy and low error rates. Conversely, LGBM shows lower R2 values (0.9061 training, 0.8809 testing) and higher MSE of 0.0747 during training and 0.0337 during testing, reflecting poorer performance. Similarly, for shrinkage prediction, XGBoost outperforms LGBM, evidenced by higher R2 (0.9887 training, 0.9975 testing) and lower MSE (0.2527 training, 0.4878 testing). The comparative statistics showed that XGBoost regularly outperformed LightGBM. The game theory-based Shapley functions revealed that temperature and product types were the most influential features of the energy consumption model. These findings illustrate the practical applicability of the XGBoost and LightGBM models in food drying operations towards optimizing drying conditions, improving product quality, and reducing energy consumption.