Size Of Dataset Research Articles

Ultra-fast Digital DPC Yielding High Spatio-temporal Resolution for Low-Dose Phase Characterization.

In the scanning transmission electron microscope, both phase imaging of beam-sensitive materials and characterization of a material's functional properties using in situ experiments are becoming more widely available. As the practicable scan speed of 4D-STEM detectors improves, so too does the temporal resolution achievable for both differential phase contrast (DPC) and ptychography. However, the read-out burden of pixelated detectors, and the size of the gigabyte to terabyte sized data sets, remain a challenge for both temporal resolution and their practical adoption. In this work, we combine ultra-fast scan coils and detector signal digitization to show that a high-fidelity DPC phase reconstruction can be achieved from an annular segmented detector. Unlike conventional analog data phase reconstructions from digitized DPC-segment images yield reliable data, even at the fastest scan speeds. Finally, dose fractionation by fast scanning and multi-framing allows for postprocess binning of frame streams to balance signal-to-noise ratio and temporal resolution for low-dose phase imaging for in situ experiments.

Computational Methods for Image Analysis in Craniofacial Development and Disease.

Observation is at the center of all biological sciences. Advances in imaging technologies are therefore essential to derive novel biological insights to better understand the complex workings of living systems. Recent high-throughput sequencing and imaging techniques are allowing researchers to simultaneously address complex molecular variations spatially and temporarily in tissues and organs. The availability of increasingly large dataset sizes has allowed for the evolution of robust deep learning models, designed to interrogate biomedical imaging data. These models are emerging as transformative tools in diagnostic medicine. Combined, these advances allow for dynamic, quantitative, and predictive observations of entire organisms and tissues. Here, we address 3 main tasks of bioimage analysis, image restoration, segmentation, and tracking and discuss new computational tools allowing for 3-dimensional spatial genomics maps. Finally, we demonstrate how these advances have been applied in studies of craniofacial development and oral disease pathogenesis.

Barley disease recognition using deep neural networks

Plant disease negatively impacts food production and quality. It is crucial to detect and recognise plant diseases correctly. Traditional approaches do not offer a rapid and comprehensive management system for detecting plant diseases. Deep learning techniques (DL) have achieved encouraging results in discriminating patterns and anomalies in visual samples. This ability provides an effective method to diagnose any plant disease symptoms automatically. However, one of the limitations of recent studies is that in-field disease detection is underexplored, so developing a model that performs well for in-field samples is necessary. The objective of this study is to develop and investigate DL techniques for in-field disease detection of barley (Hordeum vulgare L.), one of the main crops in Australia, given visual samples captured at barley trials using a consumer-grade RGB camera. Consequently, A dataset was captured from test-bed trials across multiple paddocks infected with three diseases: net form net blotch (NFNB), spot form net blotch (SFNB), and scald, in various weather conditions. The collected data, 312 images (6000 × 4000 pixels), are divided into patches of 448 × 448 pixels, which are manually annotated into four classes: no-disease, scald, NFNB and SFNB. Finally, the data was augmented using random rotation and flip to increase the dataset size. The generated barley disease dataset is then applied to several well-known pre-trained DL networks such as DenseNet, ResNet, InceptionV3, Xception, and MobileNet as the network backbone. Given limited data, these methods can be trained to detect anomalies in visual samples. The results show that MobileNet, Xception, and InceptionV3 performed well in barley disease detection. On the other hand, ResNet showed poor classification ability. Moreover, Augmenting the data improves the performance of DL networks, particularly for underperforming backbones like ResNet, and mitigates the limited data access for these data-intensive networks. The augmentation step improved MobileNet performance by approximately 6 %. MobileNet achieved the highest accuracy of 98.63 % (the average of the three diseases) in binary classification and an accuracy of 93.50 % in multi-class classification. Even though classifying SFNB and NFNB is challenging in the early stages, MobileNet achieved the minimum misclassification rate among the two diseases. The results show the efficiency of this model in diagnosing barley diseases using complex data collected from the field environment. In addition, the model is lighter and comprises fewer trainable parameters. Consequently, MobileNet is suitable for small training datasets, reducing data acquisition costs.

Time Sequence Deep Learning Model for Ubiquitous Tabular Data with Unique 3D Tensors Manipulation.

Although deep learning (DL) algorithms have been proved to be effective in diverse research domains, their application in developing models for tabular data remains limited. Models trained on tabular data demonstrate higher efficacy using traditional machine learning models than DL models, which are largely attributed to the size and structure of tabular datasets and the specific application contexts in which they are utilized. Thus, the primary objective of this paper is to propose a method to use the supremacy of Stacked Bidirectional LSTM (Long Short-Term Memory) deep learning algorithms in pattern discovery incorporating tabular data with customized 3D tensor modeling in feeding neural networks. Our findings are empirically validated using six diverse, publicly available datasets each varying in size and learning objectives. This paper proves that the proposed model based on time-sequence DL algorithms, which were generally described as inadequate when dealing with tabular data, yields satisfactory results and competes effectively with other algorithms specifically designed for tabular data. An additional benefit of this approach is its ability to preserve simplicity while ensuring fast model training also with large datasets. Even with extremely small datasets, models can be applied to achieve exceptional predictive results and fully utilize their capacity.

Improving Triage Accuracy in Prehospital Emergency Telemedicine: Scoping Review of Machine Learning-Enhanced Approaches.

Prehospital telemedicine triage systems combined with machine learning (ML) methods have the potential to improve triage accuracy and safely redirect low-acuity patients from attending the emergency department. However, research in prehospital settings is limited but needed; emergency department overcrowding and adverse patient outcomes are increasingly common. In this scoping review, we sought to characterize the existing methods for ML-enhanced telemedicine emergency triage. In order to support future research, we aimed to delineate what data sources, predictors, labels, ML models, and performance metrics were used, and in which telemedicine triage systems these methods were applied. A scoping review was conducted, querying multiple databases (MEDLINE, PubMed, Scopus, and IEEE Xplore) through February 24, 2023, to identify potential ML-enhanced methods, and for those eligible, relevant study characteristics were extracted, including prehospital triage setting, types of predictors, ground truth labeling method, ML models used, and performance metrics. Inclusion criteria were restricted to the triage of emergency telemedicine services using ML methods on an undifferentiated (disease nonspecific) population. Only primary research studies in English were considered. Furthermore, only those studies using data collected remotely (as opposed to derived from physical assessments) were included. In order to limit bias, we exclusively included articles identified through our predefined search criteria and had 3 researchers (DR, JS, and KS) independently screen the resulting studies. We conducted a narrative synthesis of findings to establish a knowledge base in this domain and identify potential gaps to be addressed in forthcoming ML-enhanced methods. A total of 165 unique records were screened for eligibility and 15 were included in the review. Most studies applied ML methods during emergency medical dispatch (7/15, 47%) or used chatbot applications (5/15, 33%). Patient demographics and health status variables were the most common predictors, with a notable absence of social variables. Frequently used ML models included support vector machines and tree-based methods. ML-enhanced models typically outperformed conventional triage algorithms, and we found a wide range of methods used to establish ground truth labels. This scoping review observed heterogeneity in dataset size, predictors, clinical setting (triage process), and reported performance metrics. Standard structured predictors, including age, sex, and comorbidities, across articles suggest the importance of these inputs; however, there was a notable absence of other potentially useful data, including medications, social variables, and health system exposure. Ground truth labeling practices should be reported in a standard fashion as the true model performance hinges on these labels. This review calls for future work to form a standardized framework, thereby supporting consistent reporting and performance comparisons across ML-enhanced prehospital triage systems.

Assessment of data‐driven modeling approaches for chromatographic separation processes

AbstractChromatographic separation processes are described by nonlinear partial differential and algebraic equations, which may result in high computational cost, hindering further online applications. To decrease the computational burden, different data‐driven modeling approaches can be implemented. In this work, we investigate different strategies of data‐driven modeling for chromatographic processes, using artificial neural networks to predict pseudo‐dynamic elution profiles, without the use of explicit temporal information. We assess the performance of the surrogates trained on different dataset sizes, achieving good predictions with a minimum of 3400 data points. Different activation functions are used and evaluated against the original high‐fidelity model, using accuracy, interpolation, and simulation time as performance metrics. Based on these metrics, the best performing data‐driven models are implemented in a process optimization framework. The results indicate that data‐driven models can capture the nonlinear profile of the process and that can be considered as reliable surrogates used to aid process development.

PharmaBench: Enhancing ADMET benchmarks with large language models

Accurately predicting ADMET (Absorption, Distribution, Metabolism, Excretion, and Toxicity) properties early in drug development is essential for selecting compounds with optimal pharmacokinetics and minimal toxicity. Existing ADMET-related benchmark sets are limited in utility due to their small dataset sizes and the lack of representation of compounds used in drug discovery projects. These shortcomings hinder their application in model building for drug discovery. To address this issue, we propose a multi-agent data mining system based on Large Language Models that effectively identifies experimental conditions within 14,401 bioassays. This approach facilitates merging entries from different sources, culminating in the creation of PharmaBench. Additionally, we have developed a data processing workflow to integrate data from various sources, resulting in 156,618 raw entries. Through this workflow, we constructed PharmaBench, a comprehensive benchmark set for ADMET properties, which comprises eleven ADMET datasets and 52,482 entries. This benchmark set is designed to serve as an open-source dataset for the development of AI models relevant to drug discovery projects.

Efficient Cryptographic Solutions for Unbalanced Private Set Intersection in Mobile Communication

Private Set Intersection (PSI) is a cryptographic method in secure multi-party computation that allows entities to identify common elements in their datasets without revealing their private data. Traditional approaches assume similar-sized datasets and equal computational power, overlooking practical imbalances. In real-world applications, dataset sizes and computational capacities often vary, particularly in the Internet of Things and mobile scenarios where device limitations restrict computational types. Traditional PSI protocols are inefficient here, as computational and communication complexities correlate with the size of larger datasets. Thus, adapting PSI protocols to these imbalances is crucial. This paper explores unbalanced PSI scenarios where one party (the receiver) has a relatively small dataset and limited computational power, while the other party (the sender) has a large amount of data and strong computational capabilities. It introduces three innovative solutions for unbalanced PSI: an unbalanced PSI protocol based on the Cuckoo filter, an unbalanced PSI protocol based on single-cloud assistance, and an unbalanced PSI protocol based on dual-cloud assistance, with each subsequent solution addressing the shortcomings of the previous one. Depending on performance and security needs, different protocols can be employed for applications such as private contact discovery.

MiLoPYP: self-supervised molecular pattern mining and particle localization in situ

Cryo-electron tomography allows the routine visualization of cellular landscapes in three dimensions at nanometer-range resolutions. When combined with single-particle tomography, it is possible to obtain near-atomic resolution structures of frequently occurring macromolecules within their native environment. Two outstanding challenges associated with cryo-electron tomography/single-particle tomography are the automatic identification and localization of proteins, tasks that are hindered by the molecular crowding inside cells, imaging distortions characteristic of cryo-electron tomography tomograms and the sheer size of tomographic datasets. Current methods suffer from low accuracy, demand extensive and time-consuming manual labeling or are limited to the detection of specific types of proteins. Here, we present MiLoPYP, a two-step dataset-specific contrastive learning-based framework that enables fast molecular pattern mining followed by accurate protein localization. MiLoPYP’s ability to effectively detect and localize a wide range of targets including globular and tubular complexes as well as large membrane proteins, will contribute to streamline and broaden the applicability of high-resolution workflows for in situ structure determination.

Does Differentially Private Synthetic Data Lead to Synthetic Discoveries?

Synthetic data have been proposed as a solution for sharing anonymized versions of sensitive biomedical datasets. Ideally, synthetic data should preserve the structure and statistical properties of the original data, while protecting the privacy of the individual subjects. Differential Privacy (DP) is currently considered the gold standard approach for balancing this trade-off. The aim of this study is to investigate how trustworthy are group differences discovered by independent sample tests from DP-synthetic data. The evaluation is carried out in terms of the tests' Type I and Type II errors. With the former, we can quantify the tests' validity, i.e., whether the probability of false discoveries is indeed below the significance level, and the latter indicates the tests' power in making real discoveries. We evaluate the Mann-Whitney U test, Student's t-test, chi-squared test, and median test on DP-synthetic data. The private synthetic datasets are generated from real-world data, including a prostate cancer dataset (n = 500) and a cardiovascular dataset (n = 70,000), as well as on bivariate and multivariate simulated data. Five different DP-synthetic data generation methods are evaluated, including two basic DP histogram release methods and MWEM, Private-PGM, and DP GAN algorithms. A large portion of the evaluation results expressed dramatically inflated Type I errors, especially at levels of ϵ ≤ 1. This result calls for caution when releasing and analyzing DP-synthetic data: low p-values may be obtained in statistical tests simply as a byproduct of the noise added to protect privacy. A DP Smoothed Histogram-based synthetic data generation method was shown to produce valid Type I error for all privacy levels tested but required a large original dataset size and a modest privacy budget (ϵ ≥ 5) in order to have reasonable Type II error levels.

STSNet: A cross-spatial resolution multi-modal remote sensing deep fusion network for high resolution land-cover segmentation

Recently, deep learning models have found extensive application in high-resolution land-cover segmentation research. However, the most current research still suffers from issues such as insufficient utilization of multi-modal information, which limits further improvement in high-resolution land-cover segmentation accuracy. Moreover, differences in the size and spatial resolution of multi-modal datasets collectively pose challenges to multi-modal land-cover segmentation. Therefore, we propose a high-resolution land-cover segmentation network (STSNet) with cross-spatial resolution spatio-temporal-spectral deep fusion. This network effectively utilizes spatio-temporal-spectral features to achieve information complementary among multi-modal data. Specifically, STSNet consists of four components: (1) A high resolution and multi-scale spatial-spectral encoder to jointly extract subtle spatial-spectral features in hyperspectral and high spatial resolution images. (2) A long-term spatio-temporal encoder formulated by spectral convolution and spatio-temporal transformer block to simultaneously delineates the spatial, temporal and spectral information in dense time series Sentinel-2 imagery. (3) A cross-resolution fusion module to alleviate the spatial resolution differences between multi-modal data and effectively leverages complementary spatio-temporal-spectral information. (4) A multi-scale decoder integrates multi-scale information from multi-modal data. We utilized airborne hyperspectral remote sensing imagery from the Shenyang region of China in 2020, with a spatial resolution of 1authors declare that they have no known competm, a spectral number of 249, and a spectral resolution ≤ 5 nm, and its Sentinel dense time-series images acquired in the same period with a spatial resolution of 10 m, a spectral number of 10, and a time-series number of 31. These datasets were combined to generate a multi-modal dataset called WHU-H2SR-MT, which is the first open accessed large-scale high spatio-temporal-spectral satellite remote sensing dataset (i.e., with >2500 image pairs sized 300 m × 300 m for each). Additionally, we employed two open-source datasets to validate the effectiveness of the proposed modules. Extensive experiments show that our multi-scale spatial-spectral encoder, spatio-temporal encoder, and cross-resolution fusion module outperform existing state-of-the-art (SOTA) algorithms in terms of overall performance on high-resolution land-cover segmentation. The new multi-modal dataset will be made available at http://irsip.whu.edu.cn/resources/resources_en_v2.php, along with the corresponding code for accessing and utilizing the dataset at https://github.com/RS-Mage/STSNet.

Recognition of Rice Species Based on Gas Chromatography-Ion Mobility Spectrometry and Deep Learning

To address the challenge of relying on complex biochemical methods for identifying rice species, a prediction model that combines gas chromatography-ion mobility spectroscopy (GC-IMS) with a convolutional neural network (CNN) was developed. The model utilizes the GC-IMS fingerprint data of each rice variety sample, and an improved CNN structure is employed to increase the recognition accuracy. First, an improved generative adversarial network based on the diffusion model (DGAN) is used for data enhancement to expand the dataset size. Then, on the basis of a residual network called ResNet50, a transfer learning method is introduced to improve the training effect of the model under the condition of a small sample. In addition, a new attention mechanism called Triplet is introduced to further highlight useful features and improve the feature extraction performance of the model. Finally, to reduce the number of model parameters and improve the efficiency of the model, a method called knowledge distillation is used to compress the model. The results of our experiments revealed that the recognition accuracy for identifying the 10 rice varieties was close to 96%; hence, the proposed model significantly outperformed traditional models such as principal component analysis and support vector machine. Furthermore, compared to the traditional CNN, our model reduced the number of parameters and number of computations by 53% and 55%, respectively, without compromising classification accuracy. The study also suggests that the combination of GC-IMS and our proposed deep learning method had better discrimination abilities for rice varieties than traditional chromatography and other spectral analysis methods and that it effectively identified rice varieties.

Optimizing Agricultural Data Analysis Techniques through AI-Powered Decision-Making Processes

The agricultural sector is undergoing a transformative paradigm shift with the integration of advanced technologies, particularly artificial intelligence (AI), to enhance data analysis techniques and streamline decision-making processes. This paper delves into the integration of advanced technologies in agriculture, focusing specifically on optimizing data analysis through artificial intelligence (AI) to strengthen decision-making processes in farming. We present a novel AI-powered model that leverages historical agricultural datasets, utilizing a comprehensive array of established machine learning algorithms to enhance the prediction and classification of agricultural data. This work provides tailored algorithm recommendations, bypassing the need to deploy and fine-tune numerous algorithms. We approximate the accuracy of suitable algorithms, highlighting those with the highest precision, thus saving time by leveraging pre-trained AI models on historical agricultural data. Our method involves three phases: collecting diverse agricultural datasets, applying multiple classifiers, and documenting their accuracy. This information is stored in a CSV file, which is then used by AI classifiers to predict the accuracy of new, unseen datasets. By evaluating feature information and various data segmentations, we recommend the configuration that achieves the highest accuracy. This approach eliminates the need for exhaustive algorithm reruns, relying on pre-trained models to estimate outcomes based on dataset characteristics. Our experimentation spans various configurations, including different training–testing splits and feature sets across multiple dataset sizes, meticulously evaluated through key performance metrics such as accuracy, precision, recall, and F-measure. The experimental results underscore the efficiency of our model, with significant improvements in predictive accuracy and resource utilization, demonstrated through comparative performance analysis against traditional methods. This paper highlights the superiority of the proposed model in its ability to systematically determine the most effective algorithm for specific agricultural data types, thus optimizing computational resources and improving the scalability of smart farming solutions. The results reveal that the proposed system can accurately predict a near-optimal machine learning algorithm and data structure for crop data with an accuracy of 89.38%, 87.61%, and 84.27% for decision tree, random forest, and random tree algorithms, respectively.

Distinct cognitive and functional connectivity features from healthy cohorts can identify clinical obsessive-compulsive disorder.

Improving diagnostic accuracy of obsessive-compulsive disorder (OCD) using models of brain imaging data is a key goal of the field, but this objective is challenging due to the limited size and phenotypic depth of clinical datasets. Leveraging the phenotypic diversity in large non-clinical datasets such as the UK Biobank (UKBB), offers a potential solution to this problem. Nevertheless, it remains unclear whether classification models trained on non-clinical populations will generalise to individuals with clinical OCD. This question is also relevant for the conceptualisation of OCD; specifically, whether the symptomology of OCD exists on a continuum from normal to pathological. Here, we examined a recently published "meta-matching" model trained on functional connectivity data from five large normative datasets (N=45,507) to predict cognitive, health and demographic variables. Specifically, we tested whether this model could classify OCD status in three independent clinical datasets (N=345). We found that the model could identify out-of-sample OCD individuals. Notably, the most predictive functional connectivity features mapped onto known cortico-striatal abnormalities in OCD and correlated with genetic brain expression maps previously implicated in the disorder. Further, the meta-matching model relied upon estimates of cognitive functions, such as cognitive flexibility and inhibition, to successfully predict OCD. These findings suggest that variability in non-clinical brain and behavioural features can discriminate clinical OCD status. These results support a dimensional and transdiagnostic conceptualisation of the brain and behavioural basis of OCD, with implications for research approaches and treatment targets.

Analysis of human ambulation as a chaotic time-series: with nonlinear dynamics tools.

The aim of the present study is to investigate the complexity and stability of human ambulation and the implications on robotic prostheses control systems. Fourteen healthy individuals participate in two experiments, the first group run at three different speeds. The second group ascended and descended stairs of a five-level building block at a self-selected speed. All participants completed the experiment with seven inertial measurement units wrapped around the lower body segments and waist. The data were analyzed to determine the fractal dimension, spectral entropy, and the Lyapunov exponent (LyE). Two methods were used to calculate the long-term LyE, first LyE calculated using the full size of data sets. And the embedding dimensions were calculated using Average Mutual Information (AMI) and the False Nearest Neighbor (FNN) algorithm was used to find the time delay. Besides, a second approach was developed to find long-term LyE where the time delay was based on the average period of the gait cycle using adaptive event-based window. The average values of spectral entropy are 0.538 and 0.575 for stairs ambulation and running, respectively. The degree of uncertainty and complexity increases with the ambulation speed. The short term LyEs for tibia orientation have the minimum range of variation when it comes to stairs ascent and descent. Using two-way analysis of variance we demonstrated the effect of the ambulation speed and type of ambulation on spectral entropy. Moreover, it was shown that the fractal dimension only changed significantly with ambulation speed.

Accelerating joint species distribution modelling with Hmsc-HPC by GPU porting.

Joint species distribution modelling (JSDM) is a widely used statistical method that analyzes combined patterns of all species in a community, linking empirical data to ecological theory and enhancing community-wide prediction tasks. However, fitting JSDMs to large datasets is often computationally demanding and time-consuming. Recent studies have introduced new statistical and machine learning techniques to provide more scalable fitting algorithms, but extending these to complex JSDM structures that account for spatial dependencies or multi-level sampling designs remains challenging. In this study, we aim to enhance JSDM scalability by leveraging high-performance computing (HPC) resources for an existing fitting method. Our work focuses on the Hmsc R-package, a widely used JSDM framework that supports the integration of various dataset types into a single comprehensive model. We developed a GPU-compatible implementation of its model-fitting algorithm using Python and the TensorFlow library. Despite these changes, our enhanced framework retains the original user interface of the Hmsc R-package. We evaluated the performance of the proposed implementation across various model configurations and dataset sizes. Our results show a significant increase in model fitting speed for most models compared to the baseline Hmsc R-package. For the largest datasets, we achieved speed-ups of over 1000 times, demonstrating the substantial potential of GPU porting for previously CPU-bound JSDM software. This advancement opens promising opportunities for better utilizing the rapidly accumulating new biodiversity data resources for inference and prediction.

Practical feature filter strategy to machine learning for small datasets in chemistry

Many potential use cases for machine learning in chemistry and materials science suffer from small dataset sizes, which demands special care for the model design in order to deliver reliable predictions. Hence, feature selection as the key determinant for dataset design is essential here. We propose a practical and efficient feature filter strategy to determine the best input feature candidates. We illustrate this strategy for the prediction of adsorption energies based on a public dataset and sublimation enthalpies using an in-house training dataset. The input of adsorption energies reduces the feature space from 12 dimensions to two and still delivers accurate results. For the sublimation enthalpies, three input configurations are filtered from 14 possible configurations with different dimensions for further productive predictions as being most relevant by using our feature filter strategy. The best extreme gradient boosting regression model possesses a good performance and is evaluated from statistical and theoretical perspectives, reaching a level of accuracy comparable to density functional theory computations and allowing for physical interpretations of the predictions. Overall, the results indicate that the feature filter strategy can help interdisciplinary scientists without rich professional AI knowledge and limited computational resources to establish a reliable small training dataset first, which may make the final machine learning model training easier and more accurate, avoiding time-consuming hyperparameter explorations and improper feature selection.

Statistical Metamodel of Liner Acoustic Impedance Based on Neural Network and Probabilistic Learning for Small Datasets

The main novelty of this paper consists of presenting a statistical artificial neural network (ANN)-based model for a robust prediction of the frequency-dependent aeroacoustic liner impedance using an aeroacoustic computational model (ACM) dataset of small size. The model, focusing on percentage of open area (POA) and sound pressure level (SPL) at a zero Mach number, takes into account uncertainties using a probabilistic formulation. The main difficulty in training an ANN-based model is the small size of the ACM dataset. The probabilistic learning carried out using the probabilistic learning on manifolds (PLoM) algorithm addresses this difficulty as it allows constructing a very large training dataset from learning the probabilistic model from a small dataset. A prior conditional probability model is presented for the PCA-based statistical reduced representation of the frequency-sampled vector of the log-resistance and reactance. It induces some statistical constraints that are not straightforwardly taken into account when training such an ANN-based model by classical optimizations methods under constraints. A second novelty of this paper consists of presenting an alternate solution that involves using conditional statistics estimated with learned realizations from PLoM. A numerical example is presented.

Enhanced Early Detection of Thyroid Abnormalities using a Hybrid Deep Learning Model: A Sequential CNN and K-Means Clustering Approach

The thyroid gland, often referred to as the butterfly gland due to its shape, is located in the neck and plays a crucial role in regulating metabolism. It is susceptible to various health conditions, including hypothyroidism, hyperthyroidism, thyroid cancer, and thyroid nodules. Early detection of these conditions is essential for accurate diagnosis and effective treatment. Detecting thyroid nodules using machine learning and deep learning techniques presents a challenging yet promising research avenue. The choice of model depends on the characteristics of the patient's thyroid data, the dataset size, and the available computational resources. Hybrid models can be employed to handle complex data more effectively. In this study, a sequential Convolutional Neural Network (CNN) model was developed due to its capability to automate feature extraction and focus on Regions-of-Interest (ROIs) for detecting thyroid abnormalities. The proposed model achieved an accuracy of 81.5%, with a precision of 97.4% and a sensitivity of 83.1%, indicating its robustness in classifying images as benign or malignant. The confusion matrix provided further performance insights. Data segmentation was enhanced using K-means clustering for its scalability and efficiency in processing large medical image datasets. Compared to traditional models, the proposed hybrid approach demonstrated a significant improvement in diagnostic accuracy and precision, achieving performance gains of approximately 15-20% over baseline methods. These advancements underscore the potential of integrating machine learning and deep learning in medical diagnostics, paving the way for more reliable and efficient diagnostic tools for healthcare professionals.

CapsRule: Explainable Deep Learning for Classifying Network Attacks.

Despite the potential deep learning (DL) algorithms have shown, their lack of transparency hinders their widespread application. Extracting if-then rules from deep neural networks is a powerful explanation method to capture nonlinear local behaviors. However, existing rule extraction methods suffer from inefficiency, incomprehensibility, infidelity, and not scaling well. Concerning security applications, they are not optimized regarding the decision boundary, data types and ranges, classification tasks, and dataset size. In this article, we propose CapsRule, an effective and efficient rule-based DL explanation method dedicated to classifying network attacks. It extracts high-fidelity rules from the feed-forward capsule network that explains how an input sample is classified. Using precomputed coupling coefficients, the training phase overlaps the rule extraction process to increase efficiency. The activation vector of a capsule can represent semantic intelligence about the attributes of the input sample. The rules extracted from CapsRule address the major concerns of network attack detection. The rules: 1) approximate the nonlinear decision boundary of the underlying data; 2) reduce the number of false positives significantly; 3) increase transparency; and 4) help find errors and noise in the data. We evaluate CapsRule on the CICDDoS2019 dataset that contains over a million of the most advanced Distributed Denial-of-Service (DDoS) attacks. The extensive evaluation shows that it generates accurate, high-fidelity, and comprehensible rules. CapsRule achieves an average accuracy of 99.0% and a false positive rate of 0.70% for reflection- and exploitation-based attacks. We verify that the learned features from the rulesets match our domain-specific knowledge. They also help find flaws in the dataset generation process and erroneous patterns caused by attack simulators.