Abstract Training a machine learning algorithm on class-imbalanced dataset can be difficult task, process that could prove even more challenging under conditions of high dimensionality. Feature extraction and data sampling are among the most popular preprocessing techniques. is used to derive richer set reduced features, while mitigate class imbalance. In this paper, we investigate these two techniques, using credit card fraud four ensemble classifiers (Random Forest, CatBoost, LightGBM,...

Abstract Acquiring labeled datasets often incurs substantial costs primarily due to the requirement of expert human intervention produce accurate and reliable class labels. In modern data landscape, an overwhelming proportion newly generated is unlabeled. This paradigm especially evident in domains such as fraud detection for credit card detection. These types have their own difficulties associated with being highly imbalanced, which poses its challenges machine learning classification. Our...

Abstract Label noise is an important data quality issue that negatively impacts machine learning algorithms. For example, label has been shown to increase the number of instances required train effective predictive models. It also model complexity and decrease interpretability. In addition, can cause classification results a learner be poor. this paper, we detect with three unsupervised learners, namely $$\textit{principal component analysis} \hbox { (PCA)}$$ <mml:math...

Abstract Fraud datasets often times lack consistent and accurate labels, are characterized by having high class imbalance where the number of fraudulent examples far fewer than those normal ones. Machine learning designed for effectively detecting fraud is an important task since behavior can have significant financial or health consequences, but presented with challenges due to availability reliable labels. This paper presents unsupervised detection method that uses iterative cleaning...

Training a machine learning algorithm from class-imbalanced dataset is an inherently challenging task. The task becomes more when compounded by high dimensionality (a number of features). Feature extraction data reduction process that transforms features into linear or non-linear combinations the original features, resulting in smaller and richer set attributes. Data sampling popular approach for addressing class imbalance. In this paper, our proposed method requires implementation feature...

One integral and necessary part of human behavior is emotion, which affects the way people communicate. Although beings can recognize interpret facial expressions, identification correct expressions continues to be a key challenging task by computer systems. The main issues stem from face's non-uniform design variations in conditions such as light, structure, posture. Several Convolutional Neural Network (CNN) approaches have been introduced for Face Emotion Recognition (FER), but these...

Typical fraud datasets lack consistent and accurate labels and, as such, are typically highly imbalanced with non-fraud examples greatly outnumbering the fraudulent ones. This presents significant challenges to machine learning researchers practitioners. Due these challenges, an effective approach in identifying data points needs handle highly-imbalanced be robust class labeling. paper introduces a novel unsupervised procedure for from without by iteratively cleaning training dataset. Our...

The COVID-19 epidemic is spreading day by day. Early diagnosis of this disease essential to provide effective preventive and therapeutic measures. This process can be used a computer-aided methodology improve accuracy. In study, new optimal method has been utilized for the COVID-19. Here, based on fuzzy <math xmlns="http://www.w3.org/1998/Math/MathML" id="M2"> <mi>C</mi> </math> -ordered means (FCOM) along with an improved version enhanced capsule network (ECN) proposed purpose. ECN mayfly...

The presence of class imbalance in machine learning datasets is a pervasive challenge that often hampers the effectiveness traditional models. In context anomaly detection, instances minority are ones most interest. To address this issue, we evaluate an unsupervised approach uses iterative cleaning process for detection on cognition data. We conduct experiments two datasets, one has large degree and other balanced. Our findings show outperforms models, namely Isolation Forest Copula-Based...

The second most prevalent age-related neurodegenerative disease is Parkinson's (PD) and Genes associated with human diseases like Parkinson are descriptive. Genome-wide association study (GWAS) used to classify the genes Parkinson’s other diseases. information of identified empowers scientists early diagnose, treat, sop Due complexities illness, identifying such a challenging task. In this article, we apply two methods feature selection choose subset that predict PD high precision in...

It is inherently challenging to train a machine learning algorithm on class-imbalanced dataset. Under conditions of high dimensionality, this training process can become even more difficult due the large number features in During preprocessing, data sampling commonly used address class imbalance and feature extraction frequently reduce dataset features. In study, we explore use these two preprocessing activities before passing four ensemble classifiers (Random Forest, CatBoost, LightGBM,...

In this paper, deep-learning-based approaches namely fine-tuning of pretrained convolutional neural networks (VGG16 and VGG19), end-to-end training a developed CNN model, have been used in order to classify X-Ray images into four different classes that include COVID-19, normal, opacity pneumonia cases. A dataset containing more than 20,000 X-ray scans was retrieved from Kaggle experiment. two-stage classification approach implemented be compared the one-shot approach. Our hypothesis model...

Privacy and security are significant issues in the field of biometric traits today's world. This research paper presents a comprehensive study that utilizes seven different deep learning models to classify Finger Knuckle Prints (FKP). The main aim this is examine efficacy fine-tuning pretrained vision adapting specific dataset being analyzed. employed include AlexNet, DensNet, EfficientNet, GoogleNet, Shallow Convolutional Neural Networks (SCNNs), ResNet50, VisionTransformer. underwent...