Search Results (14)

Traditionally, Helicobacter pylori (HP) gastritis has been diagnosed by pathologists through the examination of gastric biopsies using optical microscopy with standard hematoxylin and eosin (H&E) staining. However, with the adoption of digital pathology, the identification of HP faces certain limitations, particularly due to insufficient resolution in some scanned images. Moreover, interobserver variability has been well documented in the traditional diagnostic approach, which may further complicate consistent interpretation. In this context, deep convolutional neural network (DCNN) models are showing promising results in the automated detection of this infection in whole-slide images (WSIs). The aim of the present article is to detect the presence of HP infection from our own institutional dataset of histopathological gastric biopsy samples using different pretrained and recognized DCNN and AutoML approaches. The dataset comprises 100 H&E-stained WSIs of gastric biopsies. HP infection was confirmed previously using immunohistochemical confirmation. A total of 45,795 patches were selected for model development. InceptionV3, Resnet50, and VGG16 achieved AUC (area under the curve) values of 1. However, InceptionV3 showed superior metrics such as accuracy (97%), recall (100%), F1 score (97%), and MCC (93%). BoostedNet and AutoKeras achieved accuracy, precision, recall, specificity, and F1 scores less than 85%. The InceptionV3 model was used for external validation, and the predictions across all patches yielded a global accuracy of 78%. In conclusion, DCNN models showed stronger potential for diagnosing HP in gastric biopsies compared with the auto ML approach. However, due to variability across pathology applications, no single model is universally optimal. A problem-specific approach is essential. With growing WSI adoption, DL can improve diagnostic accuracy, reduce variability, and streamline pathology workflows using automation. Full article

Reliable streamflow forecasting is crucial for several tasks related to water-resource management, including planning reservoir operations, power generation via Hydroelectric Power Plants (HPPs), and flood mitigation, thus resulting in relevant social implications. The present study is focused on the application of Automated Machine-Learning (AutoML) models to forecast daily streamflow in the area of the upper Teles Pires River basin, located in the region of the Amazon biomes. The latter area is characterized by extensive water-resource utilization, mostly for power generation through HPPs, and it has a limited hydrological data-monitoring network. Five different AutoML models were employed to forecast the streamflow daily, i.e., auto-sklearn, Tree-based Pipeline Optimization Tool (TPOT), H2O AutoML, AutoKeras, and MLBox. The AutoML input features were set as the time-lagged streamflow and average rainfall data sourced from four rain gauge stations and one streamflow gauge station. To overcome the lack of training data, in addition to the previous features, products estimated via remote sensing were leveraged as training data, including PERSIANN, PERSIANN-CCS, PERSIANN-CDR, and PDIR-Now. The selected AutoML models proved their effectiveness in forecasting the streamflow in the considered basin. In particular, the reliability of streamflow predictions was high both in the case when training data came from rain and streamflow gauge stations and when training data were collected by the four previously mentioned estimated remote-sensing products. Moreover, the selected AutoML models showed promising results in forecasting the streamflow up to a three-day horizon, relying on the two available kinds of input features. As a final result, the present research underscores the potential of employing AutoML models for reliable streamflow forecasting, which can significantly advance water-resource planning and management within the studied geographical area. Full article

Endocrine-disrupting chemicals (EDCs) are chemicals that can interfere with homeostatic processes. They are a major concern for public health, and they can cause adverse long-term effects such as cancer, intellectual impairment, obesity, diabetes, and male infertility. The endocrine system is a complex machinery, with the estrogen (E), androgen (A), and thyroid hormone (T) modes of action being of major importance. In this context, the availability of in silico models for the rapid detection of hazardous chemicals is an effective contribution to toxicological assessments. We developed Qualitative Gene expression Activity Relationship (QGexAR) models to predict the propensities of chemically induced disruption of EAT modalities. We gathered gene expression profiles from the LINCS database tested on two cell lines, i.e., MCF7 (breast cancer) and A549 (adenocarcinomic human alveolar basal epithelial). We optimized our prediction protocol by testing different feature selection methods and classification algorithms, including CATBoost, XGBoost, Random Forest, SVM, Logistic regression, AutoKeras, TPOT, and deep learning models. For each EAT endpoint, the final prediction was made according to a consensus prediction as a function of the best model obtained for each cell line. With the available data, we were able to develop a predictive model for estrogen receptor and androgen receptor binding and thyroid hormone receptor antagonistic effects with a consensus balanced accuracy on a validation set ranging from 0.725 to 0.840. The importance of each predictive feature was further assessed to identify known genes and suggest new genes potentially involved in the mechanisms of action of EAT perturbation. Full article

This paper presents a comprehensive approach to predicting short-term (for the upcoming 2 weeks) changes in estuarine dissolved oxygen concentrations via machine learning models that integrate historical water sampling, historical and upcoming 2-week meteorological data, and river discharge and discharge metrics. Dissolved oxygen is a critical indicator of ecosystem health, and this approach is implemented for the Neuse River Estuary, North Carolina, U.S.A., which has a long history of hypoxia-related habitat degradation. Through meticulous data preprocessing and feature selection, this research evaluates the predictions of dissolved oxygen concentrations by comparing a recurrent neural network with four other models, including a Multilayer Perceptron, Long Short-Term Memory, Gradient Boosting, and AutoKeras, through sensitivity experiments. The input predictors to our prediction models include water temperature, turbidity, chlorophyll-a, aggregated river discharge, and aggregated wind based on eight directions. By emphasizing the most impactful predictors, we streamlined the model-building processes and built a hindcast system from 2015 to 2019. We found that the recurrent neural network model was most effective in predicting the dissolved oxygen concentrations, with an R² value of 0.99 at multiple stations. Different from our machine learning hindcast models that used observed upcoming meteorological and discharge data, an actual forecast system would use forecasted meteorological and discharge data. Therefore, an actual operational forecast may have lower accuracy than the hindcast, as determined by the accuracy of the predicted meteorological and discharge data. Nevertheless, our studies enhance our understanding of the factors influencing dissolved oxygen variability and set the basis for the implementation of a predictive tool for environmental monitoring and management. We also emphasized the importance of building station-specific models to improve the prediction results. Full article

Globally, over 17 million people annually die from cardiovascular diseases, with heart disease being the leading cause of mortality in the United States. The ever-increasing volume of data related to heart disease opens up possibilities for employing machine learning (ML) techniques in diagnosing and predicting heart conditions. While applying ML demands a certain level of computer science expertise—often a barrier for healthcare professionals—automated machine learning (AutoML) tools significantly lower this barrier. They enable users to construct the most effective ML models without in-depth technical knowledge. Despite their potential, there has been a lack of research comparing the performance of different AutoML tools on heart disease data. Addressing this gap, our study evaluates three AutoML tools—PyCaret, AutoGluon, and AutoKeras—against three datasets (Cleveland, Hungarian, and a combined dataset). To evaluate the efficacy of AutoML against conventional machine learning methodologies, we crafted ten machine learning models using the standard practices of exploratory data analysis (EDA), data cleansing, feature engineering, and others, utilizing the sklearn library. Our toolkit included an array of models—logistic regression, support vector machines, decision trees, random forest, and various ensemble models. Employing 5-fold cross-validation, these traditionally developed models demonstrated accuracy rates spanning from 55% to 60%. This performance is markedly inferior to that of AutoML tools, indicating the latter’s superior capability in generating predictive models. Among AutoML tools, AutoGluon emerged as the superior tool, consistently achieving accuracy rates between 78% and 86% across the datasets. PyCaret’s performance varied, with accuracy rates from 65% to 83%, indicating a dependency on the nature of the dataset. AutoKeras showed the most fluctuation in performance, with accuracies ranging from 54% to 83%. Our findings suggest that AutoML tools can simplify the generation of robust ML models that potentially surpass those crafted through traditional ML methodologies. However, we must also consider the limitations of AutoML tools and explore strategies to overcome them. The successful deployment of high-performance ML models designed via AutoML could revolutionize the treatment and prevention of heart disease globally, significantly impacting patient care. Full article

Aiming to solve the partial discharge problem caused by defects in composite insulators, most existing live detection methods are limited by the subjectivity of human judgment, the difficulty of effective quantification, and the use of a single detection method. Therefore, a composite insulator defect diagnosis model based on acoustic–electric feature fusion and a multi-scale perception multi-input of stacked auto-encoder (MMSAE) network is proposed in this paper. Initially, during the withstanding voltage experiment, the electromagnetic wave spectrometer and ultrasonic detector were used to collect and process the data of six types of composite insulator samples with artificial defects. The electromagnetic wave spectrum, ultrasonic power spectral density, and n-S map were then obtained. Then, the network architecture of MMSAE was built by integrating a stacked auto-encoder and multi-scale perception module; the feature extraction and fusion methods of the electromagnetic wave spectrum and ultrasonic signal were investigated. The proposed method was used to diagnose test samples, and the diagnostic results were compared to those obtained using a single input source and the artificial neural network (ANN) method. The results demonstrate that the detection accuracy of acoustic–electric feature fusion is greater than that of a single feature; the accuracy of the proposed method is 99.17%, which is significantly higher than the accuracy of the conventional ANN method. Finally, composite insulator defect diagnosis software based on PYQT5 and Keras was developed. Ten 500 kV aging composite insulators were used to validate the effectiveness of the proposed method and design software. Full article

The growing complexity of data derived from Industrial Internet of Things (IIoT) systems presents substantial challenges for traditional machine-learning techniques, which struggle to effectively manage the needs of predictive maintenance applications. Automated machine-learning (AutoML) techniques present a promising solution by streamlining the machine-learning process, reducing the necessity for manual hyperparameter tuning and computational resources, thereby positioning themselves as a potentially transformative innovation in the Industry 4.0 era. This research introduces two distinct models: AutoML, employing PyCaret, and Auto Deep Neural Network (AutoDNN), utilizing AutoKeras, both aimed at accurately identifying various types of faults in ball bearings. The proposed models were evaluated using the Case Western Reserve University (CWRU) bearing faults dataset, and the results showed a notable performance in terms of achieving high accuracy, recall, precision, and F1 score on the testing and validation sets. Compared to recent studies, the proposed AutoML models demonstrated superior performance, surpassing alternative approaches even when they utilized a larger number of features, thus highlighting the effectiveness of the proposed methodology. This research offers valuable insights for those interested in harnessing the potential of AutoML techniques in IIoT applications, with implications for industries such as manufacturing and energy. By automating the machine-learning process, AutoML models can help decrease the time and cost related to predictive maintenance, which is crucial for industries where unplanned downtime can lead to substantial financial losses. Full article

In this paper, we built an automated machine learning (AutoML) pipeline for structure-based learning and hyperparameter optimization purposes. The pipeline consists of three main automated stages. The first carries out the collection and preprocessing of the dataset from the Kaggle database through the Kaggle API. The second utilizes the Keras-Bayesian optimization tuning library to perform hyperparameter optimization. The third focuses on the training process of the machine learning (ML) model using the hyperparameter values estimated in the previous stage, and its evaluation is performed on the testing data by implementing the Neptune AI. The main technologies used to develop a stable and reusable machine learning pipeline are the popular Git version control system, the Google cloud virtual machine, the Jenkins server, the Docker containerization technology, and the Ngrok reverse proxy tool. The latter can securely publish the local Jenkins address as public through the internet. As such, some parts of the proposed pipeline are taken from the thematic area of machine learning operations (MLOps), resulting in a hybrid software scheme. The machine learning model was used to evaluate the pipeline, which is a multilayer perceptron (MLP) that combines typical dense, as well as polynomial, layers. The simulation results show that the proposed pipeline exhibits a reliable and accurate performance while managing to boost the network’s performance in classification tasks. Full article

Automated deep learning is promising in artificial intelligence (AI). However, a few applications of automated deep learning networks have been made in the clinical medical fields. Therefore, we studied the application of an open-source automated deep learning framework, Autokeras, for detecting smear blood images infected with malaria parasites. Autokeras is able to identify the optimal neural network to perform the classification task. Hence, the robustness of the adopted model is due to it not needing any prior knowledge from deep learning. In contrast, the traditional deep neural network methods still require more construction to identify the best convolutional neural network (CNN). The dataset used in this study consisted of 27,558 blood smear images. A comparative process proved the superiority of our proposed approach over other traditional neural networks. The evaluation results of our proposed model achieved high efficiency with impressive accuracy, reaching 95.6% when compared with previous competitive models. Full article

Diabetic retinopathy (DR) is a leading cause of blindness in middle-aged diabetic patients. Regular screening for DR using fundus imaging aids in detecting complications and delays the progression of the disease. Because manual screening takes time and is subjective, deep learning has been used to help graders. Pre-trained or brute force CNN models are used in existing DR grading CNN-based approaches that are not suited to fundus image complexity. To solve this problem, we present a method for automatically customizing CNN models based on fundus image lesions. It uses k-medoid clustering, principal component analysis (PCA), and inter-class and intra-class variations to determine the CNN model’s depth and width. The designed models are lightweight, adapted to the internal structures of fundus images, and encode the discriminative patterns of DR lesions. The technique is validated on a local dataset from King Saud University Medical City, Saudi Arabia, and two challenging Kaggle datasets: EyePACS and APTOS2019. The auto-designed models outperform well-known pre-trained CNN models such as ResNet152, DenseNet121, and ResNeSt50, as well as Google’s AutoML and Auto-Keras models based on neural architecture search (NAS). The proposed method outperforms current CNN-based DR screening methods. The proposed method can be used in various clinical settings to screen for DR and refer patients to ophthalmologists for further evaluation and treatment. Full article

Metal workpieces are indispensable in the manufacturing industry. Surface defects affect the appearance and efficiency of a workpiece and reduce the safety of manufactured products. Therefore, products must be inspected for surface defects, such as scratches, dirt, and chips. The traditional manual inspection method is time-consuming and labor-intensive, and human error is unavoidable when thousands of products require inspection. Therefore, an automated optical inspection method is often adopted. Traditional automated optical inspection algorithms are insufficient in the detection of defects on metal surfaces, but a convolutional neural network (CNN) may aid in the inspection. However, considerable time is required to select the optimal hyperparameters for a CNN through training and testing. First, we compared the ability of three CNNs, namely VGG-16, ResNet-50, and MobileNet v1, to detect defects on metal surfaces. These models were hypothetically implemented for transfer learning (TL). However, in deploying TL, the phenomenon of apparent convergence in prediction accuracy, followed by divergence in validation accuracy, may create a problem when the image pattern is not known in advance. Second, our developed automated machine-learning (AutoML) model was trained through a random search with the core layers of the network architecture of the three TL models. We developed a retraining criterion for scenarios in which the model exhibited poor training results such that a new neural network architecture and new hyperparameters could be selected for retraining when the defect accuracy criterion in the first TL was not met. Third, we used AutoKeras to execute AutoML and identify a model suitable for a metal-surface-defect dataset. The performance of TL, AutoKeras, and our designed AutoML model was compared. The results of this study were obtained using a small number of metal defect samples. Based on TL, the detection accuracy of VGG-16, ResNet-50, and MobileNet v1 was 91%, 59.00%, and 50%, respectively. Moreover, the AutoKeras model exhibited the highest accuracy of 99.83%. The accuracy of the self-designed AutoML model reached 95.50% when using a core layer module, obtained by combining the modules of VGG-16, ResNet-50, and MobileNet v1. The designed AutoML model effectively and accurately recognized defective and low-quality samples despite low training costs. The defect accuracy of the developed model was close to that of the existing AutoKeras model and thus can contribute to the development of new diagnostic technologies for smart manufacturing. Full article

Construction companies are under pressure to enhance their site safety condition, being constantly challenged by rapid technological advancements, growing public concern, and fierce competition. To enhance construction site safety, literature investigated Machine Learning (ML) approaches as risk assessment (RA) tools. However, their deployment requires knowledge for selecting, training, testing, and employing the most appropriate ML predictor. While different ML approaches are recommended by literature, their practicality at construction sites is constrained by the availability, knowledge, and experience of data scientists familiar with the construction sector. This study develops an automated ML system that automatically trains and evaluates different ML to select the most accurate ML-based construction accident severity predictors for the use of construction professionals with limited data science knowledge. A real-life accident dataset is evaluated through automated ML approaches: Auto-Sklearn, AutoKeras, and customized AutoML. The investigated AutoML approaches offer higher scalability, accuracy, and result-oriented severity insight due to their simple input requirements and automated procedures. Full article

Thermodynamic analysis of Organic Rankine Cycle (ORC) was performed in this work. The Petroleum Coke burner provided the required heat flux for the Butane Boiler. The simulation of pet-coke combustion was carried out by using Fire Dynamics Simulator software (FDS) version 5.0. Validation of the FDS calculation results was carried out by comparing the temperature of the gaseous mixture and CO₂ mole fractions to the literature. It was discovered that they are similar to those reported in the literature. An Artificial Intelligence (AI) time forecasting analysis was performed on this work. The AI algorithm was applied to the temperature and soot sensor readings. Two Python libraries were applied in order to forecast the time behaviour of the thermocouple readings: Statistical model—ARIMA (Auto-Regressive Integrated Moving Average) and KERAS—deep learning library. ARIMA is a class of model that captures a suite of different standard temporal structures in time series data. Keras is a python library applied for deep learning and runs on top of Tensor-Flow. It has been developed in order to perform deep learning models as fast and easily as possible for research and development. The model accuracy and model loss plot shows comparable performance (train and test). Butane has been employed as a working fluid in the ORC. Butane is considered one of the best pure fluids in terms of exergy efficiency. It has low specific radiative forcing (RF) compared to Ethane and Propane. Moreover, it has zero ozone depletion potential and low Global Warming Potential. It is considered flammable, highly stable and non-corrosive. The thermodynamic properties of Butane needed to evaluate the heat rate and the power were calculated by applying the ASIMPTOTE online thermodynamic calculator. It was shown that the calculated net power of the ORC cycle is similar to the net power reported in the literature (relative error of 4.8%). The proposed ORC energetic system obeys the first and second laws of thermodynamics. The thermal efficiency of the cycle is 20.4%. Full article

Automated machine learning (AutoML) has been heralded as the next wave in artificial intelligence with its promise to deliver high-performance end-to-end machine learning pipelines with minimal effort from the user. However, despite AutoML showing great promise for computer vision tasks, to the best of our knowledge, no study has used AutoML for image-based plant phenotyping. To address this gap in knowledge, we examined the application of AutoML for image-based plant phenotyping using wheat lodging assessment with unmanned aerial vehicle (UAV) imagery as an example. The performance of an open-source AutoML framework, AutoKeras, in image classification and regression tasks was compared to transfer learning using modern convolutional neural network (CNN) architectures. For image classification, which classified plot images as lodged or non-lodged, transfer learning with Xception and DenseNet-201 achieved the best classification accuracy of 93.2%, whereas AutoKeras had a 92.4% accuracy. For image regression, which predicted lodging scores from plot images, transfer learning with DenseNet-201 had the best performance (R² = 0.8303, root mean-squared error (RMSE) = 9.55, mean absolute error (MAE) = 7.03, mean absolute percentage error (MAPE) = 12.54%), followed closely by AutoKeras (R² = 0.8273, RMSE = 10.65, MAE = 8.24, MAPE = 13.87%). In both tasks, AutoKeras models had up to 40-fold faster inference times compared to the pretrained CNNs. AutoML has significant potential to enhance plant phenotyping capabilities applicable in crop breeding and precision agriculture. Full article