Multi-Class Machine Learning Classification of Fracture-Driven Interactions Using Surface Pressure Responses from Offset Wells

Automatic age and gender prediction from facial images is increasingly crucial for applications in security, marketing, and social media. Existing systems often face challenges related to accuracy, demographic generalization, and bias. This study addresses these issues by developing a deep learning-based system utilizing Convolutional Neural Networks (CNNs) for enhanced classification of age and gender. The key research gaps include limited accuracy, insufficient handling of diverse data, and model bias. The proposed approach encompasses data acquisition, preprocessing, and the design of a CNN architecture within a multi-class classification framework. Various CNN models are evaluated, incorporating transfer learning, hyperparameter optimization, and regularization techniques to improve performance. The system's effectiveness is assessed through metrics such as classification accuracy, precision, recall, and robustness across different demographic groups. Results indicate significant advancements in prediction accuracy and model generalization compared to existing methods. The technology holds practical applications in security, personalized marketing, and social networking. Challenges such as model bias and the need for diverse datasets are addressed, with future research aimed at further refining the model and expanding its applicability. This work highlights the substantial improvements deep learning offers to facial recognition technologies.

Tagged data is rapidly becoming more available on the World Wide Web. Web sites which populate tagging services offer a good way for Internet users to share their knowledge. An interesting problem is how to make tag suggestions when a new resource becomes available. In this paper, we address the issue of efficient tag suggestion. We first propose a multi-class sparse Gaussian process classification framework (SGPS) which is capable of classifying data with very few training instances. We suggest a novel prototype selection algorithm to select the best subset of points for model learning. The framework is then extended to a novel multi-class multi-label classification algorithm (MMSG) that transforms tag suggestion into the problem of multi-label ranking. Experiments on bench-mark data sets and real-world data from Del.icio.us and BibSonomy suggest that our model can greatly improve the performance of tag suggestions when compared to the state-of-the-art. Overall, our model requires linear time to train and constant time to predict per case. The memory consumption is also significantly less than traditional batch learning algorithms such as SVMs. In addition, results on tagging digital data also demonstrate that our model is capable of recommending relevant tags to images and videos by using their surrounding textual information.

This research presents an extensive point of reference for investigating the operation of several machine learning (ML) algorithms in postulating the multiclass classification problem regarding the forthcoming effects of Covid-19 on school closures. With the prompt closure of schools across the world in response to this pandemic, school-going children and teenagers are ruptured both mentally and physically. Hence, ML has come across to be a reliable component to forecast the scenario effectively. A dataset from UNESCO is trained and tested by ten supervised ML algorithms. A comprehensive analysis among the predictive ML models was executed which bought satisfactory results with regard to accuracy, precision, sensitivity, F1 score, ROC-AUC by hyper parameter optimization. In this regard, grid search cross validation (GridSearchCV) was utilized in order to obtain the optimal parameters. However, the performance of Artificial Neural Network (ANN) was also investigated and compared with the supervised ML models where ANN displayed maximum accuracy of 80.37%. After rigorous comparative analysis, Decision Tree (DT) portrayed the highest accuracy of 90.75%. Hence, it is evident that machine learning algorithm holds strong promise in forecasting the upcoming scenario of school closures due to Covid-19 and can contribute significantly in decision making for the welfare of the education system.

Selection of the most suitable drill bit type is an important task for drillers when planning for new oil and gas wells. With the advancement of intelligent predictive models, the automated selection of drill bit type is possible using earlier drilled offset wells’ data. However, real-field well data samples naturally involve an unequal distribution of data points that results in the formation of a complex imbalance multi-class classification problem during drill bit selection. In this analysis, Ensemble methods, namely Adaboost and Random Forest, have been combined with the data re-sampling techniques to provide a new approach for handling the complex drill bit selection process. Additionally, four popular machine learning techniques namely, K-nearest neighbors, naïve Bayes, multilayer perceptron, and support vector machine, are also evaluated to understand the performance degrading effects of imbalanced drilling data obtained from Norwegian wells. The comparison of results shows that the random forest with bootstrap class weighting technique has given the most impressive performance for bit type selection with testing accuracy ranges from 92% to 99%, and G-mean (0.84–0.97) in critical to normal experimental scenarios. This study provides an approach to automate the drill bit selection process over any field, which will minimize human error, time, and drilling cost.

Advanced multilateral well drilling & completion requires the application of innovative technologies while drilling to place the well in sweet spots by managing all the geological uncertainties. The smart level-4 multilateral well with dual stacked lateral was drilled by integrating advanced Real-Time geo-steering formation characterization along with geochemical, advanced gas analysis and seismic data interpretation.The Burgan reservoir consists of vertically stacked channelsands associated with geological heterogeneities along with series of fault networks connected to the aquifer at the bottom. The reservoir contains ultra-high water mobility with down hole oil viscosity of about 40cp enhances water breakthrough and requires customized ICD and ICV completions to enhance dry oil production and maximize oil recovery. Real-time geo-steering performed on Multi-lateral well by utilizing advanced technologies including High resolution Geochemical Analysis has been utilized in Real-Time to identify "geochemical proxies" and allow geochemical steering, distance to boundary tool using 4 resistivity curves with deep spacing Geosignal, At-bit measurements and density image to correct well positioning and locating faulted areas. The pre job planning had two components such as: (i) building a geosteering model based on offset well logs, geological and geophysical information and (ii) preparation of geochemical model based on XRF analysis of core chips from offset wells. The later model was calibrated through logs and utilized further to predict key rock attributes such as: (a) detailed lithological variations generally beyond the resolution of LWD logs, (b) detailed mineralogy to determine the diagenetic overprint and (c) depositional environment of different Burgan sand facies. Continuous interpretation and integration of XRF and petrophysical analysis supported by LWD data was a key factor for real-time geosteering operation. Such integrated approach also resulted in successful placing the wells with maximum reservoir contact and also was very instrumental for (i) isolation of potential trouble zones, (ii) segmentation of horizontal sections and (iii) optimization of nozzle sizes of the ICDs and hence planning of smart completion designs.

Inverter-fed induction motors are central to modern electric drive systems, widely deployed in electric vehicles (EVs), industrial automation, and renewable energy applications owing to their efficiency and robustness. However, such systems are vulnerable to critical faults, including over-current, under-voltage, and control signal anomalies, which traditional model-based or signal processing approaches struggle to detect reliably due to their sensitivity to parameter variations. To address these limitations, this study proposes a data-driven, multi-class fault classification framework using lightweight machine learning (ML) classifiers trained on a publicly available inverter-induction motor dataset comprising over 46,000 labeled time-series samples. Four operating conditions,Normal(healthy), Overcurrent, Undervoltage, and a synthetically introduced PWM Glitch, are defined using domain-informed thresholds based on signals such as DC-link voltage and three-phase stator currents. Four classifiers Random Forest, XGBoost, Gradient Boosting, and Logistic Regression are trained using an <tex xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">$80 / 20$</tex> stratified train-test split and evaluated via accuracy, weighted F1-score, confusion matrices, and ROC-AUC curves. Results indicate that the Random Forest model achieves the highest performance with 99.5% accuracy, 98.9% F1-score, and a macro-average ROC-AUC of 0.985. Feature importance analysis highlights u_dc_k(DC link voltage at time <tex xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">$\mathbf{k}$</tex>) and i_a_k(Phase A current at time <tex xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">$\mathbf{k}$</tex>) as the most critical indicators of fault conditions. This work demonstrates the effectiveness of interpretable, lightweight ML models for high-accuracy fault classification and establishes a reproducible benchmark for fault diagnosis research using open-source inverter datasets.

Electrocardiogram (ECG) classification is essential for accurately detecting and tracking heart rhythm disorders. This study proposes a multi-class ECG classification framework for identifying cardiac arrhythmias like Atrial Fibrillation (AF), Ventricular Fibrillation (VF), Normal Rhythm (NR), and Ventricular Tachycardia (VT). The ECG signals were decomposed using Variational Mode Decomposition (VMD), and higher-order statistics as well as entropy-based features were extracted from each mode. Multi-task Particle Swarm Optimization (MT-PSO) was employed to reduce redundant features and enhance the discriminative capability of the dataset. Multiple machine-learning models were evaluated, and optimized feature set led to clear performance improvements. The best results were obtained using LightGBM (ACC 0.993), HistGradientBoost (0.991), XGBoost (0.990), and ExtraTrees (0.990). Execution time also decreased for several models after optimization. Confusion-matrix and ROC analyses confirmed reliable detection across all four cardiac classes, and comparison with reported works shows that the proposed framework offers competitive or improved performance for ECG classification.

This paper proposes a novel method for multi-view face pose classification through sequential learning and sensor fusion. The basic idea is to use face images observed in visual and thermal infrared (IR) bands, with the same sampling weight in a multi-class boosting structure. The main contribution of this paper is a multi-class AdaBoost classification framework where information obtained from visual and infrared bands interactively complement each other. This is achieved by learning weak hypothesis for visual and IR band independently and then fusing the optimized hypothesis sub-ensembles. In addition, an effective feature descriptor is introduced to thermal IR images. Experiments are conducted on a visual and thermal IR image dataset containing 4844 face images in 5 different poses. Results have shown significant increase in classification rate as compared with an existing multi-class AdaBoost algorithm SAMME trained on visual or infrared images alone, as well as a simple baseline classification-fusion algorithm.

Destructive cyber-attacks and cybercriminals are increasing with the increase in IoT (Internet of Things) devices globally. This ha/s led to the need for increase in security in IoT systems. Innovative and novel techniques are used by the intruders to accomplish malicious goals effectively through cyber-attacks. An Intrusion Detection System (IDS) is used for classification of attacks in IoT networks based on anomaly detection and machine learning techniques. Inefficiency is observed in the conventional machine learning models and intrusion detection techniques as the network technologies are unpredictable. Accurate identification of various anomalies is possible with deep learning models in several research segments. The input data along with its prominent characteristics may be categorized automatically for classification and anomaly detection using convolutional neural networks (CNN). Faster computations are enabled due to the performance efficiency of CNN. For IoT networks, an intrusion detection model based on anomaly detection is designed and developed in this paper. A multiclass classification framework is created initially using a CNN model. Further, 3D CNN is used for implementation of the proposed model. Various intrusion detection datasets from IoT networks are used for validation of the proposed CNN model. Pre-trained multiclass CNN model is used for implementation of multiclass and binary classification based on transfer learning. When compared to the conventional deep learning models, the proposed multiclass and binary classification framework has attained improved F1 score, recall, precision and accuracy.

In this paper we redefine and generalize the classic k-nearest neighbors (k-NN) voting rule in a Bayesian maximum-a-posteriori (MAP) framework. Therefore, annotated examples are used for estimating pointwise class probabilities in the feature space, thus giving rise to a new instance-based classification rule. Namely, we propose to ``boost'' the classic k-NN rule by inducing a strong classifier from a combination of sparse training data, called ``prototypes''. In order to learn these prototypes, our MapBoost algorithm globally minimizes a multiclass exponential risk defined over the training data, which depends on the class probabilities estimated at sample points themselves. We tested our method for image categorization on three benchmark databases. Experimental results show that MapBoost significantly outperforms classic k-NN (up to 8%). Interestingly, due to the supervised selection of sparse prototypes and the multiclass classification framework, the accuracy improvement is obtained with a considerable computational cost reduction.

Adolescent mental health has become a critical public health concern, with increasing rates of depression, anxiety, and suicide among youth. The early identification of mental health risks is essential for timely intervention, yet traditional clinical approaches often fall short in detecting early warning signs, especially in underserved populations. This research explores the use of transformer-based deep learning models, such as BERT and RoBERTa, to analyze social media datasets for the early detection of adolescent mental health challenges. Social media platforms serve as digital diaries where young individuals often express emotions, distress, and behavioral changes, making them valuable sources for unobtrusive mental health monitoring. The study leverages annotated datasets derived from platforms like Reddit and Twitter, focusing on posts by adolescents or those discussing youth mental health. A multi-class classification framework is implemented to categorize posts into mental health risk categories, including depression, anxiety, suicidal ideation, and normal behavior. The transformer models are fine-tuned and evaluated against traditional machine learning baselines such as SVM and logistic regression, showing significantly higher accuracy, precision, and recall in detecting nuanced emotional and psychological signals. This approach aligns with the strategic goals of the Centers for Disease Control and Prevention (CDC) and the National Institutes of Health (NIH), which emphasize the importance of leveraging digital tools and AI for proactive youth mental health initiatives. Ethical considerations, including data privacy, informed consent, and bias mitigation, are addressed to ensure the responsible deployment of AI in sensitive health contexts. The findings suggest that transformer-based models can serve as effective early-warning systems, flagging at-risk individuals for further evaluation or support. By integrating this technology into school systems, mental health hotlines, or community outreach programs, stakeholders can enhance preventative care and reduce the long-term impact of adolescent mental health disorders. The study underscores the transformative potential of AI in safeguarding vulnerable youth and calls for further interdisciplinary collaboration to scale such innovations.

Classification is one of the widely used techniques to quantify and mine the rich information present in hyperspectral imagery. However, the realization of trustworthy classification still continues to be a challenging task. This is due to the presence of various uncertainties, such as incomplete a priori knowledge on the actual number of classes and noise. Especially, mismatching between the number of spectral classes and the number of information classes leads to substantial omission or commission errors in the classification. In this letter, we present a new multiclass classification framework, named supervised cascaded classifier system (SC <sup xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">2</sup> S), that addresses the abovementioned problem by providing reliable results. The SC <sup xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">2</sup> S method is a two-stage cascaded classification procedure that involves quantifying the uncertainty and then classifying the samples. In the first stage, pixels for which no reliable training samples can be found in the training stage are detected. In the second stage, pixels that are matched with the respective training distribution are classified using a supervised learning algorithm, otherwise labeled as unknown. The proposed SC <sup xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">2</sup> S framework has been implemented on eight different classification scenarios with a varying number of unknown classes (UCs) using hyperspectral and multispectral imageries. The performance of the proposed SC <sup xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">2</sup> S method is compared with other widely used classifiers with and without reject-option. The experimental results indicate that our method offers superior classification results even in the case of data sets with a large number of unknown spectral classes.

In this paper, we extend our video object recognition system to multiclass object recognition context, dealing with unbalanced data sets and comparing our resuls to state-of-the-art methods. Our approach is based on a Spatio-Temporal data representation, a dedicated kernel design and statistical learning techniques for object recognition. From video tracks made of segmented object regions in the successive frames, we extract sets of spatio-temporally coherent SIFT-based features, called Spatio-Temporal Tubes. To compare these complex tube objects, we integrate a Spatio-Temporal Tube Kernel (STTK) function into a multi-class classification framework with balancing process for unequal classes. Our approach is successfully evaluated on episodes from “Buffy, the Vampire Slayer” TV series which have been used in other works targeting same objectives. Our method proved to be more robust than dictionary based, facial feature based and key-frame based approaches. Our method is also tested on a small car database and preliminary results for car identification task illustrate its generalization potential.

In motor imagery (MI) tasks for brain computer interfaces (BCIs), the spatial covariance matrix (SCM) of electroencephalogram (EEG) signals plays a critical role in accurate classification. Given that SCMs are symmetric positive definite (SPD), Riemannian geometry is widely utilized to extract classification features. However, calculating distances between SCMs is computationally intensive due to operations like eigenvalue decomposition, and classical optimization techniques, such as gradient descent, cannot be directly applied to Riemannian manifolds, making the computation of the Riemannian mean more complex and reliant on iterative methods or approximations. In this paper, we propose a novel multiclass classification framework that integrates Riemannian geometry and neural networks to mitigate these challenges. The framework comprises two modules: a Riemannian module with multiple branches and a classification module. During training, a fusion loss function is introduced to update the branch corresponding to the true label, while other branches are updated using different loss functions along with the classification module. Comprehensive experiments on four sets of MI EEG data demonstrate the efficiency and effectiveness of the proposed model.

The measure of lumen volume on radial arteries can be used to evaluate the vessel response to different vasodilators. In this paper, we present a framework for automatic lumen segmentation in longitudinal cut images of radial artery from Intravascular ultrasound sequences. The segmentation is tackled as a classification problem where the contextual information is exploited by means of Conditional Random Fields (CRFs). A multi-class classification framework is proposed, and inference is achieved by combining binary CRFs according to the Error-Correcting-Output-Code technique. The results are validated against manually segmented sequences. Finally, the method is compared with other state-of-the-art classifiers.