Process metrics, valued for their language independence and ease of collection, have been shown to outperform product metrics in defect prediction. Among these, change entropy (Hassan, 2009) is widely used at the file level and has proven highly effective. Additionally, past research suggests that co-change patterns provide valuable insights into software quality. Building on these findings, we introduce Co-Change Graph Entropy, a novel metric that models co-changes as a graph to quantify co-change scattering.

Interpretability is vital in safety-critical domains such as defense, where understanding model behavior is crucial for building trust, ensuring accountability, and supporting decision-making. Traditional local explanation techniques, such as Local Interpretable Model-Agnostic Explanations (LIME), often neglect the causal relationships among input features. This oversight can result in misleading or spurious interpretations, particularly in complex, high-stakes environments. To address this limitation, we propose Causal LIME, an extension of LIME that incorporates causal graphs to guide the generation of perturbations in a manner consistent with the underlying data-generating mechanisms. This ensures that explanations respect the causal structure of the domain, leading to more trustworthy insights. We evaluate Causal LIME along three dimensions: (i) comparative analysis with traditional LIME, (ii) validation against permutation-based feature importance, and (iii) application to real-world military sensor data for vehicle classification. Experimental results demonstrate that Causal LIME produces more stable, causally grounded explanations, reinforcing its value in mission-critical AI applications where interpretability, reliability, and trust are paramount.

Melanoma is one of the most severe type of skin cancer caused by the DNA mutations in pigment cells called melanocytes. The current diagnosing methods involves visually examining lesion images which makes it time consuming. This makes it an active area of research which is being pursued using various deep-learning models. However, classification models that are developed using limited labeled training data might have a negative impact on the diagnostic procedure. This study presents a self-supervised learning-based model trained on dermoscopic images for automated melanoma recognition, in which feature extraction is accomplished using unlabeled data, and further fine-tuning is done using labeled data for classification. Additionally, an intermediary step is included for augmentation using Deep Convolution Generative Adversarial Networks (DCGAN) to enhance the number of unlabeled images for the pretext learning step. The trained model improved the accuracy by 4% as compared to just training using the original dataset. The proposed method can be extended to a broader range of applications in the medical area, where there is a real scarcity of data to train models.

Co-changes refer to the phenomenon wherein two or more software entities are modified together within the same commit or when they are changed to accomplish a specific task or functionality. The key to accurate co-change prediction lies in effectively predicting unseen co-changes-those occurring between entities that have not been co-changed before. However, despite considerable research on co-change patterns and prediction, there remains a significant gap in understanding unseen co-changes, including their prevalence, complexity, and predictability. We model co-changes as a graph, treating unseen co-change prediction as a link prediction task. Our method leverages file proximity measures derived from both homogeneous and heterogeneous networks, alongside other similarity measures, to predict these unseen co-changes. Analysis of 14 Apache Software Foundation projects revealed a significantly higher prevalence of unseen co-changes (up to 23x more frequent in specific projects and 7x on average) compared to recurrent co-changes. Interestingly, comparisons of co-change complexity based on file distance in the directory structure revealed no decisive differences between the two types. Our graph-based approach achieved good accuracy in predicting unseen co-changes (average AUROC of 0.84, with some projects reaching up to 0.98). Our method achieved significantly better performance than the baseline approach, demonstrating an average recall of 90% and a precision of 38%. While the precision value might seem modest, our approach achieves very high precision@k values (near 100% for 13 out of 14 projects up to$k=100$), underlining its effectiveness in real-world applications.

Abstract In recent decades, data aggregation and correlation have emerged as a significant and challenging area of research for finding useful relationships between different features of streaming data. Similarly, complex event processing (CEP) has emerged as a vital tool for aggregating and finding patterns from streaming data such as video streaming, real‐time stock trades, sensor data and more. In this context, rule‐based classifier algorithms are widely employed to discover valuable patterns from streaming data and cause‐and‐effect relationships between events. However, streaming data is dynamic and complex in nature, and it is not feasible to update the rules manually by domain experts continuously. On the contrary, dynamic data sometimes fails to adopt rules, which leads to sub optimal CEP implementation. Rule‐based CEP systems come up with their own set of challenges, which primarily includes rules adaptability for streaming IoT data. In dynamic environment, shifting patterns in data streams can impact the performance of the CEP system, as well as the correlation between events, and finding useful patterns is very much challenging. In this article, our objective is to provide a comprehensive survey for CEP using rule‐based algorithms applied across various domain and applications. We present a broad literature review using a sequential expansion of methods used for CEP, which primarily include event producers, preprocessing of events, robust rule implementation, and decision support. Additionally, we present a detailed classification of approaches used for different applications. Finally, after applying the rigorous review, it was feasible to present the state‐of‐the‐art research, which focuses on designing robust rules, application specific insights, scalable event‐driven architectures and finding the domain's trends with significant challenges and future scope.

The conventional way of respiratory coronavirus disease 2019 (COVID-19) diagnosis is reverse transcription polymerase chain reaction (RT-PCR), which is less sensitive during early stages; especially if the patient is asymptomatic, which may further cause more severe pneumonia. In this context, several deep learning models have been proposed to identify pulmonary infections using publicly available chest X-ray (CXR) image datasets for early diagnosis, better treatment and quick cure. In these datasets, the presence of less number of COVID-19 positive samples compared to other classes (normal, pneumonia and Tuberculosis) raises the challenge for unbiased learning of deep learning models. This learning problem can be considered as one-class classification problem where the target class samples are present and other classes are absent or ill-defined. All deep learning models opted class balancing techniques to solve this issue; which however should be avoided in any medical diagnosis process. Moreover, the deep learning models are also data hungry and need massive computation resources. Therefore, for quicker diagnosis, this research proposes a novel pinball loss function based one-class support vector machine (PB-OCSVM), that can work in presence of limited COVID-19 positive CXR samples (target class or class-of-interest (CoI) samples) with objectives to maximize the learning efficiency and to minimize the false predictions. The performance of the proposed model is compared with conventional OCSVM and recent deep learning models, and the experimental results prove that the proposed model outperformed state-of-the-art methods. To validate the robustness of the proposed model, experiments are also performed with noisy CXR images and UCI benchmark datasets.

Mankind has been looking up at the stars for centuries, wondering what lies in deep space, and if other civilizations like ours exist. Following this, with the significant advancement in the field of cosmology and space missions, there has been an exponential increase in the astronomical data collected by space telescopes to explore the possibility of harboring extraterrestrial life. And there is still no consensus on whether a planet is habitable, potentially habitable or inhabitable. The only habitable planet is Earth, therefore, the hypothesized exoplanets are categorized using Earth as a reference, also known as the “Earth Habitability Index” (EHI). In this regard, a number of additional metrics have been developed to categorize an exoplanet's habitability score, such as the Cobb-Douglas Habitability Score, which is a metric based on the Cobb-Douglas habitability production function (CD-HPF). Many classification-based algorithms have already been developed, but they have limitations, such as the possibility of misleading accuracy scores when applied to highly unbalanced datasets. Recently, some work has also been proposed in anomaly detection using memetic algorithms, which belong to the class of metaheuristic algorithms, but the number of feature sets used is significantly less compared to the ones impacting the habitability score of an exoplanet. In this present research, we are proposing a novel variational auto-encoder algorithm that works on a probability distribution function belonging to the class of anomaly detection and will work on a greater number of features in a significantly larger dataset. The proposed algorithm follows an unsupervised learning approach to detect anomalies in order to determine potentially habitable exoplanets. To validate the approach, the obtained results will be cross-matched with the dataset provided by the Planetary Habitability Laboratory's habitable exoplanet catalog (PHL-HEC).

Data intensive applications are one of the most critical real-time applications which are desired in most of the new-normal practices such as recommendation systems, social media analytics systems, fake news detection systems, etc. However, to deploy such solutions for real-time usage, software testing and quality assurance plays a vital role to understand the application behavior. The characteristic 4 Vs of big data adds complexities or challenges that need to be addressed for real-time applications or development. Testing of big data applications can be made efficient by designing and executing test plans; approach and strategy for all V’s of big data. This tutorial comprehensively covers both the theoretical and practical aspects of testing data-intensive applications. The tutorial discusses testing data-intensive applications built on top of modern big data frameworks such as Hadoop, Spark, Flink, NoSQL, Hive, Zookeeper, Elastic Search, Flume, and Kafka. The hands-on with MapReduce unit testing, Spark streaming testing, Kafka unit testing and related testing libraries has been covered with simple integration of testing examples and test case driven developments.

Adversarial attacks against deep learning models have gained significant attention and recent works have pro-posed explanations for the existence of adversarial examples and techniques to defend the models against these attacks. Attention in computer vision has been used to incorporate focused learning of important features and has led to improved accuracy. Recently, models with attention mechanisms have been proposed to enhance adversarial robustness. Following this context, this work aims at a general understanding of the impact of attention on adversarial robustness. This work presents a comparative study of adversarial robustness of non-attention and attention based image classification models trained on CIFAR-10, CIFAR-100 and Fashion MNIST datasets under the popular white box and black box attacks. The experimental results show that the robustness of attention based models may be dependent on the datasets used i.e. the number of classes involved in the classification. In contrast to the datasets with less number of classes, attention based models are observed to show better robustness towards classification.

Finding fuzzy frequent patterns in a quantitative database is a challenging problem of significant importance in many real-world applications. Past studies focused on mining these patterns in quantitative transactional databases by disregarding the spatiotemporal characteristics of an item in the database. This paper proposes a generic model of fuzzy frequent spatial pattern (FFSP) that may exist in a quantitative spatiotemporal database. Discovering FFSPs in a database is nontrivial and challenging due to its huge search space and high computational cost. A novel pruning technique, called neighborhood pruning, has been introduced to effectively reduce the search space and the computational cost of finding the desired itemsets. This technique facilitates the mining of FFSPs in large real-world databases practicable. We also present an efficient algorithm, called Fuzzy Frequent Spatial Pattern-Miner (FFSP-Miner), to find all desired patterns in the database. Experimental results demonstrate that FFSP-Miner is both memory and runtime efficient. Finally, we discuss the usefulness of our model with a case study on air pollution analytics.

Autonomous detection and classification of objects are admired area of research in many industrial applications. Though, humans can distinguish objects with high multi-granular similarities very easily; but for the machines, it is a very challenging task. The convolution neural networks (CNN) have illustrated efficient performance in multi-level representations of objects for classification. Conventionally, the existing deep learning models utilize the transformed features generated by the rearmost layer for training and testing. However, it is evident that this does not work well with multi-granular data, especially, in presence of deceptive similar classes (almost similar but different classes). The objective of the present research is to address the challenge of classification of deceptively similar multi-granular objects with an ensemble approach that utilizes activations from multiple layers of CNN (deep features). These multi-layer activations are further utilized to build multiple deep decision trees (known as Random forest) for classification of objects with similar appearance. The Fruits-360 dataset is utilized for evaluation of the proposed approach. With extensive trials it was observed that the proposed model outperformed over the conventional deep learning approaches.

The novel coronavirus 2019 (COVID-19) is a respiratory syndrome that resembles pneumonia. The current diagnostic procedure of COVID-19 follows reverse-transcriptase polymerase chain reaction (RT-PCR) based approach which however is less sensitive to identify the virus at the initial stage. Hence, a more robust and alternate diagnosis technique is desirable. Recently, with the release of publicly available datasets of corona positive patients comprising of computed tomography (CT) and chest X-ray (CXR) imaging; scientists, researchers and healthcare experts are contributing for faster and automated diagnosis of COVID-19 by identifying pulmonary infections using deep learning approaches to achieve better cure and treatment. These datasets have limited samples concerned with the positive COVID-19 cases, which raise the challenge for unbiased learning. Following from this context, this article presents the random oversampling and weighted class loss function approach for unbiased fine-tuned learning (transfer learning) in various state-of-the-art deep learning approaches such as baseline ResNet, Inception-v3, Inception ResNet-v2, DenseNet169, and NASNetLarge to perform binary classification (as normal and COVID-19 cases) and also multi-class classification (as COVID-19, pneumonia, and normal case) of posteroanterior CXR images. Accuracy, precision, recall, loss, and area under the curve (AUC) are utilized to evaluate the performance of the models. Considering the experimental results, the performance of each model is scenario dependent; however, NASNetLarge displayed better scores in contrast to other architectures, which is further compared with other recently proposed approaches. This article also added the visual explanation to illustrate the basis of model classification and perception of COVID-19 in CXR images.

Existing dimensionality reduction (DR) techniques such as principal component analysis (PCA) and its variants are not suitable for target class mining due to the negligence of unique statistical properties of class-of-interest (CoI) samples. Conventionally, these approaches utilize higher or lower eigenvalued principal components (PCs) for data transformation; but the higher eigenvalued PCs may split the target class, whereas lower eigenvalued PCs do not contribute significant information and wrong selection of PCs leads to performance degradation. Considering these facts, the present research offers a novel target class-guided feature extraction method. In this approach, initially, the eigendecomposition is performed on variance–covariance matrix of only the target class samples, where the higher- and lower-valued eigenvectors are rejected via statistical analysis, and the selected eigenvectors are utilized to extract the most promising feature subspace. The extracted feature-subset gives a more tighter description of the CoI with enhanced associativity among target class samples and ensures the strong separation from nontarget class samples. One-class support vector machine (OCSVM) is evaluated to validate the performance of learned features. To obtain optimized values of hyperparameters of OCSVM a novel [Formula: see text]-ary search-based autonomous method is also proposed. Exhaustive experiments with a wide variety of datasets are performed in feature-space (original and reduced) and eigenspace (obtained from original and reduced features) to validate the performance of the proposed approach in terms of accuracy, precision, specificity and sensitivity.

With the increasing applications of deep learning in biomedical image analysis, in this article we introduce an inception U-Net architecture for automating nuclei detection in microscopy cell images of varying size and modality to help unlock faster cures, inspired from Kaggle Data Science Bowl Challenge 2018 (KDSB18). This study follows from the fact that most of the analysis requires nuclei detection as the starting phase for getting an insight into the underlying biological process and further diagnosis. The proposed architecture consists of a switch normalization layer, convolution layers, and inception layers (concatenated 1x1, 3x3, and 5x5 convolution and the hybrid of a max and Hartley spectral pooling layer) connected in the U-Net fashion for generating the image masks. This article also illustrates the model perception of image masks using activation maximization and filter map visualization techniques. A novel objective function segmentation loss is proposed based on the binary cross entropy, dice coefficient, and intersection over union loss functions. The intersection over union score, loss value, and pixel accuracy metrics evaluate the model over the KDSB18 dataset. The proposed inception U-Net architecture exhibits quite significant results as compared to the original U-Net and recent U-Net++ architecture.

Drug counterfeiting is a serious and increasing critical issue worldwide, which puts the health of consumers and the general population at risk. WHO estimates that the world's counterfeit drug market make billions of dollars annually. It is due to the inadequate supply chain in the developing countries that one in the drug is counterfeit. People buying such medicines not only waste their money but also put their life and health in danger. A drug's ownership changes from manufacturer to distributor and then to the pharmacist before reaching the customer. The manufacturers don't know how their drug is being used. At the same time, consumers don't know whether the drug has come from the rightful source. It is due to the counterfeit drug in the supply chain manufacturing companies and countries face huge economic loss. The RFID technology has been effective for over a decade providing anti-counterfeit measures to the supply chain. However, the RFID tags cannot be guaranteed genuineness, since the tags can be cloned easily. In this paper, we propose how blockchain can be utilized to not only prevent drug counterfeiting but also add traceability, security, and visibility to the pharmaceutical supply chain. For this purpose, a permissioned blockchain is designed to which only trusted parties can join the network and push transactions to the blockchain.