Unknown Patterns Research Articles

Artificial intelligence contributes to the creative transformation and innovative development of traditional Chinese culture

In recent years, artificial intelligence (AI) has emerged as a transformative force in various fields, including the arts and culture. This is particularly evident in the context of traditional Chinese culture, where AI has become a powerful tool in its creative transformation and innovative development. With its advanced capabilities in data processing and generating new ideas, AI is not only helping to preserve the rich heritage of Chinese culture but is also playing a crucial role in its evolution. This study aims to delve into how AI is reshaping the traditional elements of Chinese culture, such as calligraphy, Chinese paintings and traditional artworks, and assess its impact on both conservation and modern reinterpretation. We also examine real-world applications and projects that utilize AI technologies, such as machine learning, natural language processing, and computer vision. Our findings indicate that AI's contribution to traditional Chinese culture is multifaceted. One of the key areas where AI has made a significant impact is in the preservation and restoration of cultural artifacts. AI algorithms have demonstrated remarkable proficiency in analyzing large datasets of historical texts and artworks, uncovering previously unknown patterns and facilitating the restoration of ancient texts and relics. The integration of artificial intelligence into the realm of traditional Chinese culture signifies a pivotal moment in its history. AI's role extends beyond mere preservation; it is a catalyst for innovation, fostering new forms of artistic expression and promoting a dynamic cross-cultural exchange. As AI technology continues to evolve, it is expected to further revolutionize the way we interact with and understand traditional Chinese culture, opening up new avenues for creative exploration and cultural dialogue. This study underscores the potential of AI as a tool for cultural enrichment and highlights the exciting prospects for future developments in this area.

Spermidine synthase promotes liver cancer progression in a paracrine manner by altering the macrophage immunometabolic state.

Understanding the molecular mechanisms of adaptive regulation in the tumor microenvironment is crucial for precision therapy in hepatocellular carcinoma (HCC). We hypothesized that cargo proteins carried by extracellular vesicles (EVs) released in a hypoxic microenvironment might promote HCC progression by remodeling tumor-associated macrophages (TAMs). EV protein analysis by label-free proteomics mass spectrometry of HCC cell lines of different tumor grades was performed. The promotional effect if spermidine synthase（SRM） on M2 polarized TAMs was further investigated using various biological approaches. SRM expression was positively correlated with liver cancer progression in HCC cell lines, liver cancer samples, and nude mouse models. In a mouse model, SRM expression was positively correlated with TAM infiltration and liver cancer progression. Pan-cancer dataset analysis confirmed that SRM overexpression in HCC tumors is correlated with poor patient prognosis. However, a hypoxic microenvironment is an internal driving factor for exosomal SRM that participates in microenvironmental modifications. Moreover, we defined a hitherto unknown pattern of microenvironmental crosstalk involving SRM in EVs, whereby macrophages complete the phenotypic fate of M2 tumor-associated macrophages through SRM uptake. SRM regulation within the immune microenvironment is metabolically driven. By upregulating spermidine, which serves as a substrate for eIF5A hypusination, excessive oxidative phosphorylation (OXPHOS) assembly is achieved. This, in turn, leads to the expression of immunosuppressive marker molecules and ultimately promotes liver cancer progression. SRM, which is enriched in the EVs of HCC cells under hypoxic conditions, acts as a potent regulator linking polyamine and energy metabolism in TAMs, thereby promoting liver cancer progression.

Evolution of different adaptations for pollen uptake from flowers of Heliotropium (Boraginaceae) with narrow corolla tube in pollen wasps of the genus Celonites (Hymenoptera: Vespidae: Masarinae)

Abstract The evolution of flower-visiting behaviour in pollen wasps remains poorly investigated. The females of oligolectic Celonites species show two fundamentally different behavioural patterns for pollen uptake from Heliotropium flowers with a narrow corolla tube. They remove pollen from the concealed anthers either with their forelegs or with their proboscis. Single-frame video analysis revealed that pollen collection with the proboscis consists of a hitherto unknown, unique, and stereotypic behavioural pattern. The first phylogenetic reconstruction of Celonites based on ultra-conserved elements demonstrates that this highly complex behaviour is presumably homologous and evolved only once. It probably derived from accidental consumption of pollen adhering to the proboscis during nectar uptake and is associated with morphological adaptations of the foretarsi and mouthparts. Pollen uptake with the forelegs evolved independently. Morphological adaptations to this type of pollen uptake are the elongation of the forelegs and a tarsal pollen brush formed by hooked setae. These results confirm for pollen wasps a view that has been developed in studies of bee–flower relationships, namely that pollen collection, contrary to nectar collection, is a specific and stereotypic behaviour that often requires morphological and behavioural adaptations, possibly underlying the high rate of specialization observed in pollen-collecting Hymenoptera.

Clinical and Genetic Characteristics of 40 Patients with Non-Muscle Myosin Heavy Chain 9-Related Disease (MYH9-RD) Misdiagnosed as ITP: A Retrospective Analysis in China.

Myosin Heavy Chain 9-related diseases (MYH9-RD) are rare autosomal dominant platelet disorders characterised by macrothrombocytopaenia and leukocyte inclusion bodies. They can manifest with non-haematological complications, including deafness, nephropathy, or cataracts. Due to its rarity and its similar clinical presentation with immune thrombocytopaenia (ITP), MYH9-RD is often misdiagnosed as ITP, leading to inappropriate treatment and delayed management of complications. The study aimed to evaluate clinical, therapeutic, and genetic aspects of patients with MYH9-RD misdiagnosed with ITP, comparing differences between Chinese paediatric and adult cases of this condition. This multi-centre retrospective study included data obtained from Chinese patients diagnosed with MYH9-RD between January 2014 and December 2023 at five centres. Adults exhibited significantly longer median misdiagnosis (9 years vs. 0.2 years, p < 0.001) and treatment durations (1.5 years vs. 0.1 years, p < 0.001) than children. Non-haematological manifestations were exclusive to adults (10/21). All patients received inappropriate ITP treatments, with adults receiving more different treatments. Genetic analysis revealed 21 spontaneous mutations (52.5%), 12 familial mutations, and 7 mutations of unknown inheritance patterns. Two novel mutations (p.G1517V and p.K1674Q) were identified. Patients with p.R702C mutation demonstrated early-stage kidney injury and hearing loss. Adult patients with MYH9-RD face elevated risks of misdiagnosis, prolonged inappropriate treatment, and non-haematological complications than paediatric patients. Enhanced awareness, consideration of mean platelet volume, family history, and genetic screening are crucial for accurate MYH9-RD diagnosis and management. The prevalence of spontaneous mutations and identified genotype-phenotype correlations warrant further investigation in the Chinese population.

Stock selection with intuitionistic fuzzy combined compromise solutions

In the dynamic landscape of stock market investments, decision-making processes are often challenged by uncertainty and imprecision inherent in the financial data. It is important to evaluate the suitability of each stock based on multiple criteria, including market trends, financial indicators, and risk factors. Various approaches can be utilized to determine best option based on different parameters. The ambiguity and uncertainty around a phenomena are handled by fuzzy sets. Intuitionistic fuzzy (IF) set is an extension of fuzzy set used for managing uncertainty in more complex situations when fuzzy sets are unable to produce reliable results. This paper outlines a technique in IF environment to solve the problem of selection of stocks based on financial and other factors. A new IF knowledge measure (IFKM) is proposed to measure the amount of knowledge linked to IF sets. Reliability and utility of introduced knowledge measure is tested using some numerical examples. A new IF accuracy measure (IFAM) is proposed based on the suggested knowledge measure. Proposed IFAM is used to find the patterns similarity of unknown pattern with given pattern. Besides this, a new score function is proposed to compare the IF numbers which can get around the drawbacks of the current scoring functions. A modified Combined compromise solution (CoCoSo) technique is provided to choose the best stock by using the suggested scoring function and the IF accuracy measure. Lastly, to prove the effectiveness of the suggested method, a comparative analysis carries out with the other existing methods.

High-throughput capture of transcription factor-driven epigenome dynamics using PHILO ChIP-seq.

Assessing the dynamics of chromatin features and transcription factor (TF) binding at scale remains a significant challenge in plants. Here, we present PHILO (Plant HIgh-throughput LOw input) ChIP-seq, a high-throughput ChIP-seq platform that enables the cost-effective and extensive capture of TF binding and genome-wide distributions of histone modifications. The PHILO ChIP-seq pipeline is adaptable to many plant species, requires very little starting material (1mg), and provides the option to use MNase (micrococcal nuclease) for chromatin fragmentation. By employing H3K9ac PHILO ChIP-seq on eight Arabidopsis thaliana jasmonic acid (JA) pathway mutants, with the simultaneous processing of over 100 samples, we not only recapitulated but also expanded the current understanding of the intricate interplay between the master TFs MYC2/3/4 and various chromatin regulators. Additionally, our analyses brought to light previously unknown histone acetylation patterns within the regulatory regions of MYC2 target genes in Arabidopsis, which is also conserved in tomato (Solanum lycopersicum). In summary, our PHILO ChIP-seq platform demonstrates its high effectiveness in investigating TF binding and chromatin dynamics on a large scale in plants, paving the way for the cost-efficient realization of complex experimental setups.

Enhanced detection of unknown defect patterns on wafer bin maps based on an open-set recognition approach

It is crucial to detect and classify defect patterns on wafers in semiconductor-manufacturing processes for wafer-quality management and prompt analysis of defect causes. In recent years, continuous technological innovation and advancements in semiconductor-industry processes have led to an increase in unknown defect patterns, which must be detected and classified. However, detection of unknown defect patterns is difficult due to complex reasons, such as training on non-existent defect classes, closed datasets owing to industrial security, and labeling large volumes of manufacturing data. Owing to these challenges, methods for detecting unknown defect patterns in an actual semiconductor-manufacturing environment primarily rely on qualitative indicators, such as intuition and experience of engineers. To overcome these problems, this study proposes a methodology based on open-set recognition to accurately detect unknown defect patterns. This methodology begins with two preprocessing steps: constrained mean filtering (C-mean filtering); and Radon transform to diminish noise and efficiently extract features from wafer-bin maps. This study then develops an entropy-estimation one-class support vector machine (EEOC-SVM), which accounts for the uncertainty in the one-class SVM classification results. EEOC-SVM computes entropy-uncertainty scores based on the distance between decision boundaries and samples and then reclassifies uncertain samples using a weighted sum of uncertainties for each class. This method can effectively detect unknown defect patterns. The proposed method achieves a detection performance of over 98 % for various defect classes based on experiments conducted with new defect patterns occurring in actual semiconductor-manufacturing environments. These results confirm that the proposed method is an effective tool for detecting and addressing unknown defect patterns.

APPLIED MIXED KERNEL AND FOURIER SERIES MODELLING IN NONPARAMETRIC REGRESSION

There are three nonparametric regression approaches, namely, parametric, nonparametric and semi-parametric regression. Nonparametric regression allows the response variable to follow a different curve from one predictor variable to another. In paired data, the components of predictor variable and the response variable are assumed to follow unknown data patterns, so that they can be approximated by a kernel based regression model and Fourier series. The base components are approximated by kernel functions and Fourier series functions. Error is assumed to be normally distributed with zero mean and constant variance. The aim of this research is to make a kernel-based nonparametric regression model and Fourier series on poverty data in the Province of Bali. The research phase method begins with the introduction of the kernel based nonparametric regression model and the Fourier series. The next step is to study the estimation of the regression curve. The function estimation results are highly dependent on the bandwidth, smoothing and oscillation parameters. The resulting model gives a value of R2=0.6278, means that the used variables can explain the model by 62.78 percent. From the obtained modeling re-sults, the Open Unemployment Rate has a positive effect on the percentage of poor people in Bali. Keywords: Nonparametric Regression; Modeling; Kernel; Fourier Series.

Challenges in Mathematics Arising from Theoretical and Computational Chemistry

The combination of theoretical chemistry and mathematics represents a significant advancement in science since it has yielded insights into the enigmas surrounding atoms and molecules. The advancement of technology has greatly benefited drug development, leading to significant advancements in both medicine and materials research. This summary discusses innovative methodologies and computational challenges in the expansive domain of theoretical and computational chemistry. Despite the computational challenges associated with the Schrodinger equation and Density Functional Theory (DFT), quantum mechanics offers a basic understanding of the behaviour of molecules and atoms. While MD simulations may not handle long-range interactions and complex configuration spaces, they nonetheless excel at capturing various time scales of molecular behaviour. Nevertheless, the process of transforming abstract notions into computer models for the purpose of solving MD simulations continues to be difficult; statistical mechanics offers a theoretical framework for understanding MD simulations. Data-driven methodologies and machine learning have significantly transformed the perspective of computational chemistry by enabling the observation and comprehension of very complex chemical interactions. We must also address concerns about interpretability, reliability in limited data streams, and system transferability. These challenges expedite the advancement of novel materials, product designs, and pharmaceuticals by leveraging machine learning and integrating it with specialised domain expertise. We used available data from many reputable databases, spanning the time period from 2010 to 2024. It is imperative to prioritise the advancement of multiscale modelling techniques and hybrid quantum-classical tools in order to achieve sufficient mastery over chemical phenomena. Another potential benefit is the use of novel mathematical models to uncover previously unknown patterns and connections in genetic data. The most effective way to tackle mathematical problems is by employing a straightforward approach and engaging in collaboration with experts from other fields. By adopting this approach, we can effectively address significant societal issues and propel advancements in chemistry. Keywords: mathematics, computational, chemistry, theoretical, challenges, quantum mechanics

Decoding transaminase motifs: Tracing the unknown patterns for enhancing the accuracy of computational screening methodologies

Transaminases, enzymes known for their amino group transfer capabilities, encompass four distinct subfamilies: D-alanine transaminase (DATA), L-selective Branched chain aminotransferase (BCAT), and 4-amino-4-deoxychorismate lyase (ADCL) and R-selective aminotransferase (RATA). RATA enzymes are particularly valuable in biocatalysis for synthesizing chiral amines and resolving racemic mixtures, yet their identification in sequence databases is challenging due to the lack of robust motif-based screening methods. Constructing a sequence dataset of transaminases, and categorizing them to various subfamilies, the conserved motifs are screened over the experimentally known ones, and the novel motifs are explored. Phylogenetic clustering of these subfamilies and structural localization of the identified motifs on the Alphafold-predicted protein models of the representative sequences validate their functional importance. For the ADCL, BCAT, DATA, and RATA datasets, we identified 5, 7, 10, and 2 novel motifs, with 3, 5, 7, and 2 motifs localized on secondary structures, confirming their structural importance. Furthermore, the analysis revealed 1, 3, 2, and 1 unique residue patterns of 293-KxxxR-297; 336-KxxxxY-341, 379-ExxxxNxF-386, and 453-ExFxxGT-459; 187-HxxRL-191, and 284-DxRWxxCDIK-293; and 191-HxxRL-195, integrating of which in the known computational tools would improve their accuracy. The conserved residue pattern or motif-based computational approach for robustly screening the transaminases holds promise for unveiling the novel RATA enzymes, facilitating their exploitation in biocatalytic applications.

Enhancing IoT Security Using GA-HDLAD: A Hybrid Deep Learning Approach for Anomaly Detection

The adoption and use of the Internet of Things (IoT) have increased rapidly over recent years, and cyber threats in IoT devices have also become more common. Thus, the development of a system that can effectively identify malicious attacks and reduce security threats in IoT devices has become a topic of great importance. One of the most serious threats comes from botnets, which commonly attack IoT devices by interrupting the networks required for the devices to run. There are a number of methods that can be used to improve security by identifying unknown patterns in IoT networks, including deep learning and machine learning approaches. In this study, an algorithm named the genetic algorithm with hybrid deep learning-based anomaly detection (GA-HDLAD) is developed, with the aim of improving security by identifying botnets within the IoT environment. The GA-HDLAD technique addresses the problem of high dimensionality by using a genetic algorithm during feature selection. Hybrid deep learning is used to detect botnets; the approach is a combination of recurrent neural networks (RNNs), feature extraction techniques (FETs), and attention concepts. Botnet attacks commonly involve complex patterns that the hybrid deep learning (HDL) method can detect. Moreover, the use of FETs in the model ensures that features can be effectively extracted from spatial data, while temporal dependencies are captured by RNNs. Simulated annealing (SA) is utilized to select the hyperparameters necessary for the HDL approach. In this study, the GA-HDLAD system is experimentally assessed using a benchmark botnet dataset, and the findings reveal that the system provides superior results in comparison to existing detection methods.

Utilizing echocardiography and unsupervised machine learning for heart failure risk identification

BackgroundGlobal longitudinal strain (GLS) is recognized as a powerful predictor of heart failure (HF). However, the entire strain curve may entail important prognostic information regarding HF risk that might be undiscovered by only focusing on the peak strain value. ObjectiveThe hypothesis of the present study was, that analysis of the entire strain curve using unsupervised machine learning (uML) would reveal novel ventricular deformation patterns capable of predicting incident HF independently of GLS. MethodsLongitudinal strain curves from 3710 subjects from the general population without prevalent HF were analyzed using uML. ResultsMean age was 56 years and 43 % were male. During a median follow-up of 5.3 years, 92 subjects (2.5 %) developed HF. The uML algorithm generated a hierarchical clustering tree (HCT) resulting in 10 different clusters. Generally, the strain curves displayed reduced early diastolic strain to peak-strain ratio with an increasing incidence rate of HF. In multivariable Cox regressions, cluster 9 was significantly associated with increased risk of HF when compared to cluster 2–5, and 7–8 [For cluster 3: HR 8.95, 95 %CI: 2.08;38.48, P = 0.003] even though the subjects of cluster 9 were younger, displayed healthier clinical baseline characteristics, and only had slightly reduced GLS. The mean strain curve of cluster 9 displayed an early systolic lengthening followed by a late and reduced contraction specifically related to the basal lateral segment. ConclusionThe unsupervised machine learning algorithm identified unknown strain patterns beyond GLS presumably related to increased risk of HF.

Expanding Short-Term Pedestrian Volume Counts to Average Annual Daily Counts at Intersections: Temporal Pattern Indices and a Comparison of Statistical Land Use and Simple Threshold Methods

Quantifying pedestrian exposure for intersections is important for developing pedestrian-centric strategies. Exposure is typically expressed as the annual-average daily pedestrian traffic (AADPT). AADPT can be directly calculated when continuous counts (CCs) are available. However, CCs are rarely available for all intersections, requiring estimation methods. This work explores the steps involved in expanding pedestrian short-term counts (STCs) to AADPT using factor groups. Day-of-week-of-month expansion factors were used to expand 8-h STCs. The first step consists of grouping sites with similar pedestrian activity patterns (factor groups). Five factor groups were identified, and three pedestrian temporal pattern indicators were proposed to characterize them. The thresholds of these indicators for grouping sites with CCs are available in this work. The second step involves the challenge of associating sites where only STCs are available (unknown traffic patterns) with a particular factor group. Multinomial logistic regression (MLR) was developed to link factor groups to land use, socioeconomic, and operational attributes. Explanatory variables that represent commercial and industrial areas and the number of schools and transit stops were found to be significant. Simpler methods were also introduced to identify factor groups strongly affected by schools. The last step assesses the expansion of STCs to AADPT using the proposed approaches. The results showed improved performance when compared with averaging expansion factors across all sites. It was also demonstrated that, in certain cases, simple approaches that rely exclusively on the identification of sites that are highly affected by schools present similar performance to more complex approaches, such as MLR.

Efficient degradation representation learning network for remote sensing image super-resolution

The advancements in convolutional neural networks have led to significant progress in image super-resolution (SR) techniques. Nevertheless, it is crucial to acknowledge that current SR methods operate under the assumption of bicubic downsampling as a degradation factor in low-resolution (LR) images and train models accordingly. However, this approach does not account for the unknown degradation patterns present in real-world scenes. To address this problem, we propose an efficient degradation representation learning network (EDRLN). Specifically, we adopt a contrast learning approach, which enables the model to distinguish and learn various degradation representations in realistic images to obtain critical degradation information. We also introduce streamlined and efficient pixel attention to strengthen the feature extraction capability of the model. In addition, we optimize our model with mutual affine convolution layers instead of ordinary convolution layers to make it more lightweight while minimizing performance loss. Experimental results on remote sensing and benchmark datasets show that our proposed EDRLN exhibits good performance for different degradation scenarios, while the lightweight version minimizes the performance loss as much as possible. The Code will be available at: https://github.com/Leilei11111/EDRLN.

IrCLIP-RNP and Re-CLIP reveal patterns of dynamic protein associations on RNA.

RNA binding proteins ( RBPs ) control varied processes, including RNA splicing, stability, transport, and translation 1-3 . Dysfunctional RNA-RBP interactions contribute to the pathogenesis of human disease 1,4,5 , however, characterizing the nature and dynamics of multiprotein assemblies on RNA has been challenging. To address this, non-isotopic ligation-based ultraviolet crosslinking immunoprecipitation 6 was combined with mass spectrometry ( irCLIP-RNP ) to identify RNA-dependent associated proteins ( RDAPs ) co-bound to RNA with any RBP of interest. irCLIP-RNP defined landscapes of multimeric protein assemblies on RNA, uncovering previously unknown patterns of RBP-RNA associations, including cell-type-selective combinatorial relationships between RDAPs and primary RBPs. irCLIP-RNP also defined dynamic RDAP remodeling in response to epidermal growth factor ( EGF ), uncovering EGF-induced recruitment of UPF1 adjacent to HNRNPC to effect splicing surveillance of cell proliferation mRNAs. To identify the RNAs simultaneously co-bound by multiple studied RBPs, a sequential immunoprecipitation irCLIP ( Re-CLIP ) method was also developed. Re-CLIP confirmed binding relationships seen in irCLIP-RNP and detected simultaneous HNRNPC and UPF1 co-binding on RND3 and DDX3X mRNAs. irCLIP-RNP and Re-CLIP provide a framework to identify and characterize dynamic RNA-protein assemblies in living cells.

Implementasi Algoritma K-Means Guna Pengelompokkan Data Penjualan Berdasarkan Pembelian

Information technology can help solve problems faced by humans by facilitating performance. Information technology and information systems are difficult to separate in the business world. Data mining is the core of the KDD process, which involves inferring algorithms that explore data, developing models and finding previously unknown patterns. KDD is often referred to as knowledge discovery in databases. The KDD process generally consists of 5 stages, namely data selection, pre-processing/cleaning, transformation, data mining and interpretation/evaluation. K-Means is a clustering algorithm in data mining to be able to produce groups of large amounts of data with a point-based partition method with fast and efficient computing time. Clustering is the process of dividing objects from a data set into several homogeneous clusters. The main purpose of the cluster method is to group a number of data/objects into clusters (groups) so that each cluster will contain data that is as similar as possible. This study aims to provide real solutions to UD. Martua in order to know which items are selling well and which items are not selling well so that the object can know which items need to be added to the stock and which items need to be reduced. The method used in this study is the K-Means method with stages, namely data selection, pre-processing, data transformation, information extraction and evaluation of results. The data consists of 30 item data, there are 8 as members of C1 and are best-selling items and 22 as members of C2 and are not selling items. The conclusion that can be obtained from this study is that the K-Means method can group items at UD. Martua. This study shows that the implementation of the K-Means method with the support of the RapidMiner application is effective in grouping item data at UD. Martua.

Proteome-Scale Tissue Mapping Using Mass Spectrometry Based on Label-Free and Multiplexed Workflows

Multiplexed bimolecular profiling of tissue microenvironment, or spatial omics, can provide deep insight into cellular compositions and interactions in healthy and diseased tissues. Proteome-scale tissue mapping, which aims to unbiasedly visualize all the proteins in a whole tissue section or region of interest, has attracted significant interest because it holds great potential to directly reveal diagnostic biomarkers and therapeutic targets. While many approaches are available, however, proteome mapping still exhibits significant technical challenges in both protein coverage and analytical throughput. Since many of these existing challenges are associated with mass spectrometry-based protein identification and quantification, we performed a detailed benchmarking study of three protein quantification methods for spatial proteome mapping, including label-free, TMT-MS2, and TMT-MS3. Our study indicates label-free method provided the deepest coverages of ∼3500 proteins at a spatial resolution of 50μm and the highest quantification dynamic range, while TMT-MS2 method holds great benefit in mapping throughput at >125 pixels per day. The evaluation also indicates both label-free and TMT-MS2 provides robust protein quantifications in identifying differentially abundant proteins and spatially covariable clusters. In the study of pancreatic islet microenvironment, we demonstrated deep proteome mapping not only enables the identification of protein markers specific to different cell types, but more importantly, it also reveals unknown or hidden protein patterns by spatial coexpression analysis.

MCGnet with multi-level gated unit transformations: A time series prediction-based model for unknown pattern abnormality detection

The task of anomaly detection in data patterns remains challenging due to the diverse of fault patterns, which often render existing models ineffective when encountered with unknown or unseen data patterns in the training data. Furthermore, insufficient samples, the presence of chaotic behavior, unbalanced anomaly patterns are potential limiting factors in the practical application. While prediction error-based approaches attempt to address these challenges, the discriminability of abnormal patterns in existing scenarios is not yet sufficient, thus limiting the ability to identify abnormal phenomena. To overcome these limitations, we proposed a novel model called MCGnet. The backbone architecture of MCGnet employed a novel design of continuous multi-layer circular convolution transformations. This design enhances the expressive capability of the model by expanding the differences among data patterns. Additionally, we introduced in parallel the newly proposed temporal-frequency gated Gaussian unit, which generates richer feature representations. Subsequently, the obtained feature representations were fed into an explicit causal network to learn complex patterns of data along temporal dependencies. The MCGnet not only solves high-precision time series prediction problems but also significantly improves the discriminability between normal and abnormal patterns. Since the proposed model is accurately modeled using only a small amount of normal data, it bypasses the limiting factors such as insufficient fault samples and sample imbalance, and is able to effectively deal with various unknown data patterns. The accuracy of the proposed method was verified through simulation data (Duffing system) and its potential for engineering applications was demonstrated through real-world data (rotor system).

Epileptic network identification: insights from Dynamic Mode Decomposition of sEEG data.

For medically-refractory epilepsy patients, stereoelectroencephalography (sEEG) is a surgical method using intracranial recordings to identify brain networks participating in early seizure organization and propagation (i.e., the epileptogenic zone, EZ). If identified, surgical EZ treatment via resection, ablation or neuromodulation can lead to seizure-freedom. To date, quantification of sEEG data, including its visualization and interpretation, remains a clinical and computational challenge. Given elusiveness of physical laws or governing equations modelling complex brain dynamics, data science offers unique insight into identifying unknown patterns within high dimensional sEEG data. We apply here an unsupervised data-driven algorithm, Dynamic Mode Decomposition (DMD), to sEEG recordings from five focal epilepsy patients (three with temporal lobe, and two with cingulate epilepsy), who underwent subsequent resective or ablative surgery and became seizure free. DMD obtains a linear approximation of nonlinear data dynamics, generating coherent structures ("modes") defining important signal features, used to extract frequencies, growth rates and spatial structures. DMD was adapted to produce Dynamic Modal Maps (DMMs) across frequency sub-bands, capturing onset and evolution of epileptiform dynamics in sEEG data. Additionally, we developed a static estimate of EZ-localized electrode contacts, termed the Higher-Frequency Mode-based Norm Index (MNI). DMM and MNI maps for representative patient seizures were validated against clinical sEEG results and seizure-free outcomes following surgery. DMD was most informative at higher frequencies, i.e. gamma (including high-gamma) and beta range, successfully identifying EZ contacts. Combined interpretation of DMM/MNI plots best identified spatiotemporal evolution of mode-specific network changes, with strong concordance to sEEG results and outcomes across all five patients. The method identified network attenuation in other contacts not implicated in the EZ. This is the first application of DMD to sEEG data analysis, supporting integration of neuroengineering, mathematical and machine learning methods into traditional workflows for sEEG review and epilepsy surgical decision-making.

How interdisciplinary is it? A new method for quantifying interdisciplinarity in student teams & examination of a large project-based program

ABSTRACT The European Higher Education Area and the US National Academies of Science, Engineering, and Medicine have called on higher education to support interdisciplinary learning. The National Academies recommend evaluating interdisciplinary education against relevant criteria, but beyond listing student majors, institutions cannot easily quantify interdisciplinarity in education. The Rao-Stirling diversity index, which incorporates measures of cognitive distance, is used to quantify the interdisciplinarity of research. This study extends the use of the Rao-Stirling diversity index to the context of higher education to: (1) determine whether it provides a more nuanced measure of interdisciplinarity than number of majors, and (2) to examine patterns in interdisciplinarity across 79 student teams in Georgia Tech’s Vertically Integrated Projects Program (N = 1218). The analysis employed cognitive distances developed in a previous study. Distributions of diversity indexes and enrollment by major were examined across the program, along with visualizations for 4 teams. Results show the Rao-Stirling diversity index provided a more comprehensive measure of team interdisciplinarity, and it indicated higher interdisciplinarity in more teams and in comparatively smaller teams than the number of majors would imply. Indexes across the program were examined graphically and explored through regression. Results indicated that higher interdisciplinarity was associated with smaller team size, implying a possible need for additional support for these teams. The utility of the index in quantifying the interdisciplinarity of teams and in revealing a previously unknown pattern confirm its usefulness within the context of higher education. Limitations, implications for research and practice, and future directions are discussed.