Pattern Feature Research Articles

A FRAMEWORK TO ANALYZE OS SYSTEMS ARTIFACTS FROM LINUX MACHINES

Memory forensics plays a pivotal role in digital investigations, providing crucial insights into the activities and artifacts of an operating system. This paper explores the application of deep learning techniques in the domain of memory forensics within the Linux environment. Linux-based systems are widely used in various contexts, including servers, embedded devices, and desktops, making memory analysis in this ecosystem of paramount importance. Traditional memory forensics techniques have relied on manual analysis, which is often time-consuming and error prone. Deep learning, a subfield of machine learning, has demonstrated remarkable capabilities in pattern recognition and feature extraction tasks. In response, this paper presents a novel framework that automates and improves memory analysis through deep learning. Key components of this framework include data collection, preprocessing, feature extraction, and model selection. We introduce a unique dataset specifically curated for Linux memory forensics, facilitating the development and evaluation of deep learning models. Our experimental results demonstrate the efficacy of using a ResNet-50 model for detecting and classifying malware from memory dumps, achieving a detection rate of 98.75% and an accuracy rate of 89% in classifying malware types. Additionally, we acknowledged the challenges and limitations of applying deep learning in memory forensics, such as model interpretability and data privacy concerns. Future research directions are discussed, including real-time memory analysis integration and techniques for handling encrypted and compressed memory data.

Design of a differentiable L-1 norm for pattern recognition and machine learning

In various applications of pattern recognition, feature selection, and machine learning, L-1 norm is used as either an objective function or a regularizer. Mathematically, L-1 norm has unique characteristics that make it attractive in machine learning, feature selection, optimization, and regression. Computationally, however, L-1 norm presents a hurdle as it is non-differentiable, making the process of finding a solution difficult. Existing approach therefore relies on numerical approaches. In this work we designed an L-1 norm that is differentiable and, thus, has an analytical solution. The differentiable L-1 norm removes the absolute sign in the conventional definition and is everywhere differentiable. The new L-1 norm is almost everywhere linear, a desirable feature that is also present in the conventional L-1 norm. The only limitation of the new L-1 norm is that near zero, its behavior is not linear, hence we consider the new L-1 norm quasi-linear. Being differentiable, the new L-1 norm and its quasi-linear variation make them amenable to analytic solutions. Hence, it can facilitate the development and implementation of many algorithms involving L-1 norm. Our tests validate the capability of the new L-1 norm in various applications.

TPat: Transition pattern feature extraction based Parkinson’s disorder detection using FNIRS signals

Background and ObjectiveParkinson’s Disease (PD) is one of the most commonly observed neurodegenerative disorders worldwide. Many researchers have utilized machine learning (ML) models to detect PD and understand its underlying causes automatically. In this research, our primary objective is to automatically detect PD and extract meaningful results using the proposed ML model. Materials and MethodsIn this study, an FNIRS dataset collected from PD patients and control participants under three conditions—(i) rest, (ii) walking, and (iii) finger tapping—was utilized. A new explainable feature engineering (XFE) model was proposed to detect PD and automatically extract meaningful information under these conditions. The XFE model consists of four main phases: (i) feature extraction using the proposed channel transformation and transition pattern (TPat), (ii) feature selection employing cumulative weighted neighborhood component analysis (CWNCA), (iii) classification using the k-nearest neighbors (kNN) classifier, and (iv) channel network extraction to obtain explainable results. ResultsThe suggested TPat-based XFE model was applied to the FNIRS dataset. This dataset included three distinct cases. Our model achieved over 94% classification accuracy using leave-one-subject-out cross-validation (LOSO CV) and 100% classification accuracy using 10-fold cross-validation. Additionally, channel transitions for each case were identified and discussed. ConclusionsBased on the results and findings, the proposed model demonstrated high accuracy in FNIRS signal classification and provided explainable results. In this regard, the presented TPat-based XFE model contributed significantly to both ML and neuroscience.

Fingerprint Pattern Feature Extraction for Loop Fingerprint Pattern Identification by Zhang-Suen and Stentiford Thinning Method

Every fingerprint that a human possesses must have a fingerprint pattern and must be unique. Each pattern has a ridge pattern that will not change as long as there is no change due to an accident or injury. This ridge pattern can be used as biometric recognition. Based on the ridge pattern owned, the fingerprint pattern is divided into 4 patterns, including the tented arch pattern, the whorl pattern, the loop pattern (ulnar loop and radial loop). The stage of fingerprint pattern recognition begins by looking for minutiae (termination and bifurcation) points on the fingerprint using the Crossing Number (CN) method. Before searching for CN, the fingerprint image must be processed using the Otsu and Zhang-Suen or Stentiford Thinning methods. Thinning is used to facilitate the extraction of minutiae in fingerprint images. The last stage in obtaining the minutiae point is classified by the Linear Discriminant Analysis (LDA) method. The results obtained from the test data were 30 images, 24 images could be classified correctly, and 4 images could not be recognized. The accuracy rate of the fingerprint pattern identification system is 80%.

Alzheimer’s disease (AD) classification using MRI: A deep ensemble model with modified local pattern feature set

The Alzheimer disease (AD) is a neurologic brain condition, which affects the cells in the brain and eventually renders a patient incapable of performing routine daily tasks. Due to the outstanding spatial clarity, high access, and strong contrast, MRI has been utilized in analyses pertaining to AD. This work develops an AD classification model using MRI images. Here, preprocessing is done by the Gabor filter. Subsequently, the Improved U-net segmentation model is employed for image segmentation. The features extracted comprises of modified LGXP features, LTP features, and LBP features as well. Finally, the Deep ensemble classifier (DEC) model is proposed for AD classification which combines classifiers such as RNN, DBN, and Deep Maxout Network (DMN). For enhancing the efficiency for classification of AD, the optimal weight of DMN is adjusted using the Self Customized BWO (SC-BWO) model. The outputs from DEC are averaged and the final result is obtained. Finally, the analysis of dice, Jaccard scores is performed to show the betterment of the SC-BWO scheme.

Concurrent control chart pattern recognition in manufacturing processes based on zero-shot learning

In real industrial settings, collecting and labeling concurrent abnormal control chart pattern (CCP) samples are challenging, thereby hindering the effectiveness of current CCP recognition (CCPR) methods. This paper introduces zero-shot learning into quality control, proposing an intelligent model for recognizing zero-shot concurrent CCPs (C-CCPs). A multiscale ordinal pattern (OP) feature considering data sequential relationship is proposed. Drawing from expert knowledge, an attribute description space (ADS) is established to infer from single CCPs to C-CCPs. An ADS is embedded between features and labels, and the attribute classifier associates the features and attributes of CCPs. Experimental results demonstrate an accuracy of 98.73% for 11 unseen C-CCPs and an overall accuracy of 98.89% for all 19 CCPs, without C-CCP samples in training. Compared with other features, the multiscale OP feature has the best recognition effect on unseen C-CCPs.

The Framework of Quantifying Biomarkers of OCT and OCTA Images in Retinal Diseases.

Despite the significant advancements facilitated by previous research in introducing a plethora of retinal biomarkers, there is a lack of research addressing the clinical need for quantifying different biomarkers and prioritizing their importance for guiding clinical decision making in the context of retinal diseases. To address this issue, our study introduces a novel framework for quantifying biomarkers derived from optical coherence tomography (OCT) and optical coherence tomography angiography (OCTA) images in retinal diseases. We extract 452 feature parameters from five feature types, including local binary patterns (LBP) features of OCT and OCTA, capillary and large vessel features, and the foveal avascular zone (FAZ) feature. Leveraging this extensive feature set, we construct a classification model using a statistically relevant p value for feature selection to predict retinal diseases. We obtain a high accuracy of 0.912 and F1-score of 0.906 in the task of disease classification using this framework. We find that OCT and OCTA's LBP features provide a significant contribution of 77.12% to the significance of biomarkers in predicting retinal diseases, suggesting their potential as latent indicators for clinical diagnosis. This study employs a quantitative analysis framework to identify potential biomarkers for retinal diseases in OCT and OCTA images. Our findings suggest that LBP parameters, skewness and kurtosis values of capillary, the maximum, mean, median, and standard deviation of large vessel, as well as the eccentricity, compactness, flatness, and anisotropy index of FAZ, may serve as significant indicators of retinal conditions.

Individualized generation of women's prototype based on the classification of body shape

The traditional prototype method only relies on the basic dimensions of the human body such as bust girth and back length to make the prototype pattern, ignoring the differences of human body details dimensions. In order to improve the fit of garments, the research on intelligent generation technology of individualized garment patterns has become one of the hot spots in the field of clothing. Based on the 3D body scanning technology, we proposed a method for generating prototype pattern based on body shape classification. The proposed method takes as inputs the body parameters (width, thickness and angle), while the output of the method is one of the four pattern generation rules-"Y" shape, "V" shape, "I" shape, or "X" shape. 207 subjects were divided into four categories based on the body angle parameters, namely "Y" shape, "V" shape, "I" shape and "X" shape, and a shape recognition model was built. The pattern database is required to obtain pattern generation rules by 3D point cloud reconstruction and flattening. Combined with the human body shape parameters, the mathematical formulas of the feature points on the pattern landmarks are built. The accuracy of body shape recognition model is 97.4%. The error analysis of the predicted pattern parameters shows that 90% of the pattern feature points have a goodness of fit above 0.7. In the 5 main landmarks, the proportion of pattern within the absolute error range is more than 80%, indicating that the prediction effect of the pattern is good. This method can be applied to the process of automatic pattern generation system based 2D measurement to improve work efficiency.

Self-consistent sharp interface theory of active condensate dynamics

Biomolecular condensates help organize the cell cytoplasm and nucleoplasm into spatial compartments with different chemical compositions. A key feature of such compositional patterning is the local enrichment of enzymatically active biomolecules which, after transient binding via molecular interactions, catalyze reactions among their substrates. Thereby, biomolecular condensates provide a spatial template for nonuniform concentration profiles of substrates. In turn, the concentration profiles of substrates, and their molecular interactions with enzymes, drive enzyme fluxes which can enable novel nonequilibrium dynamics. To analyze this generic class of systems, with a current focus on self-propelled droplet motion, we here develop a self-consistent sharp interface theory. In our theory, we diverge from the usual bottom-up approach, which involves calculating the dynamics of concentration profiles based on a given chemical potential gradient. Instead, reminiscent of control theory, we take the reverse approach by deriving the chemical potential profile and enzyme fluxes required to maintain a desired condensate form and dynamics. The chemical potential profile and currents of enzymes come with a corresponding power dissipation rate, which allows us to derive a thermodynamic consistency criterion for the passive part of the system (here, reciprocal enzyme-enzyme interactions). As a first-use case of our theory, we study the role of reciprocal interactions, where the transport of substrates due to reactions and diffusion is, in part, compensated by redistribution due to molecular interactions. More generally, our theory applies to mass-conserved active matter systems with moving phase boundaries. Published by the American Physical Society 2024

Discriminant Input Processing Scheme for Self-Assisted Intelligent Healthcare Systems.

Modern technology and analysis of emotions play a crucial role in enabling intelligent healthcare systems to provide diagnostics and self-assistance services based on observation. However, precise data predictions and computational models are critical for these systems to perform their jobs effectively. Traditionally, healthcare monitoring has been the primary emphasis. However, there were a couple of negatives, including the pattern feature generating the method's scalability and reliability, which was tested with different data sources. This paper delves into the Discriminant Input Processing Scheme (DIPS), a crucial instrument for resolving challenges. Data-segmentation-based complex processing techniques allow DIPS to merge many emotion analysis streams. The DIPS recommendation engine uses segmented data characteristics to sift through inputs from the emotion stream for patterns. The recommendation is more accurate and flexible since DIPS uses transfer learning to identify similar data across different streams. With transfer learning, this study can be sure that the previous recommendations and data properties will be available in future data streams, making the most of them. Data utilization ratio, approximation, accuracy, and false rate are some of the metrics used to assess the effectiveness of the advised approach. Self-assisted intelligent healthcare systems that use emotion-based analysis and state-of-the-art technology are crucial when managing healthcare. This study improves healthcare management's accuracy and efficiency using computational models like DIPS to guarantee accurate data forecasts and recommendations.

Attention-transfer-based path loss prediction in asymmetric massive MIMO IoT systems

The asymmetric massive multiple-input–multiple-output (MIMO) array improves system capacity and provides wide-area coverage for the Internet of Things (IoT). In this paper, we propose a novel attention-based model for path loss (PL) prediction in asymmetric massive MIMO IoT systems. To represent the propagation characteristics, the propagation image that considers the detailed environment, beamwidth pattern, and propagation-statistics feature is designed. Benefiting from the shuffle attention computation, the proposed model, termed a shuffle-attention-based convolutional neural network (SAN), can effectively extract the detailed features of the propagation scenario from the image. Besides, we design the beamwidth-scenario transfer learning (BWSTL) algorithm to assist the SAN model in predicting PL in the new asymmetric massive MIMO IoT systems, where the beamwidth configuration and propagation scenario are different. It is shown that the proposed model outperforms the empirical model and other state-of-the-art artificial intelligence-based models. Aided by the BWSTL algorithm, the SAN model can be transferred to new propagation conditions with limited samples, which is beneficial to the fast deployment in the new asymmetric massive MIMO IoT systems.

CC-GAIN: Clustering and classification-based generative adversarial imputation network for missing electricity consumption data imputation

The widespread use of data across various fields has made missing data imputation technology a crucial tool. High-quality data is essential for effective energy management in smart grid environments, but essential data may be absent during collection. Despite the development of various imputation techniques for electricity consumption data, previous studies have made limited efforts to address the distinctive characteristics of this domain adequately. To overcome this limitation, a high-performance imputation model must effectively leverage time-series and pattern features of the data. This study proposes a novel missing data imputation model based on unsupervised clustering and classification-based generative adversarial imputation network(CC-GAIN), which excels in pattern classification and feature extraction. The CC-GAIN model demonstrates superior performance across all types and rates of missing data, outperforming alternative models.

AI-driven discovery of blood xenobiotic biomarkers in neovascular age-related macular degeneration using iterative random forests.

To investigate the xenobiotic profiles of patients with neovascular age-related macular degeneration (nAMD) undergoing anti-vascular endothelial growth factor (anti-VEGF) intravitreal therapy (IVT) to identify biomarkers indicative of clinical phenotypes through advanced AI methodologies. In this cross-sectional observational study, we analyzed 156 peripheral blood xenobiotic features in a cohort of 46 nAMD patients stratified by choroidal neovascularization (CNV) control under anti-VEGF IVT. We employed Liquid Chromatography-Tandem Mass Spectrometry (LC-MS/MS) for measurement and leveraged an AI-driven iterative Random Forests (iRF) approach for robust pattern recognition and feature selection, aligning molecular profiles with clinical phenotypes. AI-augmented iRF models effectively refined the metabolite spectrum by discarding non-predictive elements. Perfluorooctanesulfonate (PFOS) and Ethyl β-glucopyranoside were identified as significant biomarkers through this process, associated with various clinically relevant phenotypes. Unlike single metabolite classes, drug metabolites were distinctly correlated with subretinal fluid presence. This study underscores the enhanced capability of AI, particularly iRF, in dissecting complex metabolomic data to elucidate the xenobiotic landscape of nAMD and environmental impact on the disease. The preliminary biomarkers discovered offer promising directions for personalized treatment strategies, although further validation in broader cohorts is essential for clinical application.

LSTM-QDCNN: long short-term memory and quantum dilated convolutional neural network enabled occlusion percentage prediction

ABSTRACT Occlusion detection is vital in longer video sequences of real-time scenarios, as deformable objects generally occlude themselves or be occluded by exterior objects eventually. Occlusion can affect accuracy of the entire model significantly, because optical flow may reveal random attributes in occluded as well as neighbourhood areas. Occlusion is a challenging issue in real-world visual object tracking. Several techniques learn imprecise appearance of target while it becomes occluded by another object in the scene. Occlusion is a complicated issue that causes target loss in tracking procedure. Here, long short-term memory fused quantum dilated convolutional neural network (LSTM-QDCNN) is presented for occlusion percentage prediction. Initially, video frames are extracted from considered input video. The extracted frame is pre-processed by adaptive fuzzy filtering, and object localisation is accomplished by employing sparse fuzzy c-means with local optimal oriented pattern feature. Thereafter, occlusion category detection is conducted by deep Q-network. Lastly, occlusion percentage prediction is done by utilising LSTM-QDCNN. Additionally, LSTM-QDCNN is designed by merging deep LSTM (DLSTM) with quantum dilated convolutional neural network (QDCNN). Furthermore LSTM-QDCNN achieved maximum multiple-object tracking precision of 0.892 and minimum tracking distance of 3.714. In addition, LSTM-QDCNN achieved tracking number about 18.

Application of graph-curvature features in computer-aided diagnosis for histopathological image identification of gastric cancer

BackgroundHistopathology diagnosis is often regarded as the final diagnostic method for malignant tumors; however, it has some drawbacks. This study explored a computer-aided diagnostic method that can be used to identify benign and malignant gastric cancer using histopathological images. MethodsThe most suitable process was selected through multiple experiments by comparing multiple methods and features for classification. First, the U-net was applied to segment the image. Next, the nucleus was extracted from the segmented image, and the minimum spanning tree (MST) diagram structure that can capture the topological information was drawn. The third step was to extract the graph-curvature features of the histopathological image according to the MST image. Finally, by inputting the graph-curvature features into the classifier, the recognition results for benign or malignant cancer can be obtained. ResultsDuring the experiment, we used various methods for comparison. In the image segmentation stage, U-net, watershed algorithm, and Otsu threshold segmentation methods were used. We found that the U-net method, combined with multiple indicators, was the most suitable for segmentation of histopathological images. In the feature extraction stage, in addition to extracting graph-edge and graph-curvature features, several basic image features were extracted, including the red, green and blue feature, gray-level co-occurrence matrix feature, histogram of oriented gradient feature, and local binary pattern feature. In the classifier design stage, we experimented with various methods, such as support vector machine (SVM), random forest, artificial neural network, K nearest neighbors, VGG-16, and inception-V3. Through comparison and analysis, it was found that classification results with an accuracy of 98.57% can be obtained by inputting the graph-curvature feature into the SVM classifier. ConclusionThis study created a unique feature, the graph-curvature feature, based on the MST to represent and analyze histopathological images. This graph-based feature could be used to identify benign and malignant cells in histopathological images and assist pathologists in diagnosis.

Modeling terminal Pleistocene and Holocene forager population increase and environmental change in the Central Namib desert, Namibia

This paper examines the interplay of environmental change and human demographic shifts among late Pleistocene and Holocene populations in the hyper-arid context of the Central Namib Gravel plains, Western Namibia. This paper applies a structural equation modeling (SEM) approach to arid-zone hunter-gatherer mobility patterns and settlement systems based on variability among ethnographically and historically observed groups. It then applies this model in examining issues related to the Middle-to-Later Stone Age transition at the sites of Erb Tanks and the Mirabib rock shelter. This model suggests that, among modern arid-zone foragers, population density is by far the strongest factor influencing mobility and settlement systems, with higher population densities favoring longer residential site usages and shorter residential moves within smaller territories. In contrast, environmental variables having to do with annual rainfall, rainfall seasonality, and effective temperature affect mobility and settlement systems in ambiguous and statistically insignificant ways. This paper closes by arguing that regional-scale population increases signaled by various features of Later Stone Age archaeological patterning led to local population expansions in the Central Namib Desert, and that this in turn caused a range of recognizable cultural shifts that were strongly linked with adaptations to aridity.

Recent Advances in Positive Photoresists: Mechanisms and Fabrication.

Photoresists are fundamental materials in photolithography and are crucial for precise patterning in microelectronic devices, MEMS, and nanostructures. This paper provides an in-depth review of recent advancements in positive photoresist research and development, focusing on discussion regarding the underlying mechanisms governing their behavior, exploring innovative fabrication techniques, and highlighting the advantages of the photoresist classes discussed. The paper begins by discussing the need for the development of new photoresist technologies, highlighting issues associated with adopting extreme ultraviolet photolithography and addressing these challenges through the development of advanced positive-tone resist materials with improved patterning features, resolution, and sensitivity. Subsequently, it discusses the working mechanisms and synthesis methods of different types and subtypes of photoresists, starting from non-chemically amplified, organic, and inorganic-organic hybrid photoresists and progressing to dry film resists, with an emphasis on the upsides of each. The paper concludes by discussing how future research in the field of lithography-prioritizing concerns related to environmental impacts, improved photoresist material and properties, and utilization of advanced quantum technology-can assist with revolutionizing lithography techniques.

Species-Agnostic Patterned Animal Re-identification by Aggregating Deep Local Features

Access to large image volumes through camera traps and crowdsourcing provides novel possibilities for animal monitoring and conservation. It calls for automatic methods for analysis, in particular, when re-identifying individual animals from the images. Most existing re-identification methods rely on either hand-crafted local features or end-to-end learning of fur pattern similarity. The former does not need labeled training data, while the latter, although very data-hungry typically outperforms the former when enough training data is available. We propose a novel re-identification pipeline that combines the strengths of both approaches by utilizing modern learnable local features and feature aggregation. This creates representative pattern feature embeddings that provide high re-identification accuracy while allowing us to apply the method to small datasets by using pre-trained feature descriptors. We report a comprehensive comparison of different modern local features and demonstrate the advantages of the proposed pipeline on two very different species.

Adaptive Threshold and Weighted Frequency Domain Histogram of Local Binary Patterns

Wire ropes are crucial load-bearing components in mining conveyance equipment, and machine vision is one of the methods used to assess the surface damage condition of wire ropes. In response to the light-sensitive nature of local binary patterns, which leads to issues such as differing feature values for similar textures and susceptibility to the influence of excessively large or small pixels within local windows, hindering the accurate reflection of window structure information and exacerbating the introduction of considerable feature noise, an investigation is conducted. To enhance the gradient structural information among pixels within local pixel window, an adaptive threshold binary pattern feature operator is proposed. This operator utilizes the mean and variance within the local window to balance the central pixel value, thereby enhancing the interconnection among neighboring pixels. To perform feature selection on block histograms, a block-weighted approach is employed. This approach utilizes the concept of block weighting and employs correlation coefficients to preprocess feature vectors, thereby enhancing classification accuracy. The algorithm experiments were conducted on a dataset of mine wire ropes. The results indicate that the improved local binary pattern significantly enhances the classification accuracy of the wire rope dataset, achieving an accuracy of 97.3%.

Enhancing fall prediction in the elderly people using LBP features and transfer learning model

In an era where the detection and prevention of falls are crucial for the well-being of elderly and vulnerable individuals, achieving high accuracy in identifying such incidents is of utmost importance. This study introduces a modified NASNet, a novel transformer learning model designed for the classification of fall and not fall people based on local binary patterns (LBP) features. The modified NASNet model demonstrates high performance, achieving an accuracy, precision, recall, and F1 score of 99% with LBP features in differentiating fall from not fall. To assess the performance of NASNet, this research work conducted a comparative analysis with other prominent deep learning, and transfer learning models, as well as state-of-the-art machine learning frameworks. The findings indicate NASNet's enhanced performance in fall detection, highlighting its potential for application in real-time fall detection systems, contributing to a safer environment for individuals at risk. This research sets the stage for enhanced healthcare and assistance for vulnerable populations, addressing a critical concern in the healthcare sector.