检索结果-内蒙古大学图书馆

Oblique and rotation double random forest

学校读者我要写书评

暂无评论

arXiv 2021年

作者： Ganaie, M.A. Tanveer, M. Suganthan, P.N. Snasel, V. Department of Mathematics Indian Institute of Technology Indore Simrol Indore453552 India School of Electrical & Electronic Engineering Nanyang Technological University Singapore KINDI Center for Computing Research College of Engineering Qatar University Qatar Department of Computer Science VŠB - Technical University of Ostrava Czech Republic

Random Forest is an ensemble of decision trees based on the bagging and random subspace concepts. As suggested by Breiman, the strength of unstable learners and the diversity among them are the ensemble models’ core strength. In this paper, we propose two approaches known as oblique and rotation double random forests. In the first approach, we propose rotation based double random forest. In rotation based double random forests, transformation or rotation of the feature space is generated at each node. At each node different random feature subspace is chosen for evaluation, hence the transformation at each node is different. Different transformations result in better diversity among the base learners and hence, better generalization performance. With the double random forest as base learner, the data at each node is transformed via two different transformations namely, principal component analysis and linear discriminant analysis. In the second approach, we propose oblique double random forest. Decision trees in random forest and double random forest are univariate, and this results in the generation of axis parallel split which fails to capture the geometric structure of the data. Also, the standard random forest may not grow sufficiently large decision trees resulting in suboptimal performance. To capture the geometric properties and to grow the decision trees of sufficient depth, we propose oblique double random forest. The oblique double random forest models are multivariate decision trees. At each nonleaf node, multisurface proximal support vector machine generates the optimal plane for better generalization performance. Also, different regularization techniques (Tikhonov regularisation, axis-parallel split regularisation, Null space regularisation) are employed for tackling the small sample size problems in the decision trees of oblique double random forest. The proposed ensembles of decision trees produce trees with bigger size compared to the standard ensembl

关键词： Decision trees

The Role of Masks in Mitigating Viral Spread on Networks

学校读者我要写书评

暂无评论

arXiv 2021年

作者： Tian, Yurun Sridhar, Anirudh Wu, Chai Wah Levin, Simon A. Carley, Kathleen M. Poor, H. Vincent Yağan, Osman Department of Electrical and Computer Engineering Carnegie Mellon University PittsburghPA15213 United States Department of Electrical and Computer Engineering Princeton University PrincetonNJ08544 United States Thomas J. Watson Research Center IBM Yorktown HeightsNY10598 United States Department of Ecology & Evolutionary Biology Princeton University PrincetonNJ08544 United States Software and Societal Systems School of Computer Science Carnegie Mellon University PittsburghPA15213 United States

Masks have remained an important mitigation strategy in the fight against COVID-19 due to their ability to prevent the transmission of respiratory droplets between individuals. In this work, we provide a comprehensive quantitative analysis of the impact of mask-wearing. To this end, we propose a novel agent-based model of viral spread on networks where agents may either wear no mask or wear one of several types of masks with different properties (e.g., cloth or surgical). We derive analytical expressions for three key epidemiological quantities: the probability of emergence, the epidemic threshold, and the expected epidemic size. In particular, we show how the aforementioned quantities depend on the structure of the contact network, viral transmission dynamics, and the distribution of the different types of masks within the population. Through extensive simulations, we then investigate the impact of different allocations of masks within the population and trade-offs between the outward efficiency and inward efficiency of the masks. Interestingly, we find that masks with high outward efficiency and low inward efficiency are most useful for controlling the spread in the early stages of an epidemic, while masks with high inward efficiency but low outward efficiency are most useful in reducing the size of an already large spread. Lastly, we study whether degree-based mask allocation is more effective in reducing the probability of epidemic as well as epidemic size compared to random allocation. The result echoes the previous findings that mitigation strategies should differ based on the stage of the spreading process, focusing on source control before the epidemic emerges and on self-protection after the emergence. Copyright © 2021, The Authors. All rights reserved.

关键词： Efficiency

Current advancements and prospects of enzymatic and non-enzymatic electrochemical glucose sensors

学校读者我要写书评

暂无评论

International Journal of Biological Macromolecules 2023年第PT 2期253卷 126680页

作者： Govindaraj, Muthukumar Srivastava, Ananya Muthukumaran, Magesh Kumar Tsai, Pei-Chien Lin, Yuan-Chung Raja, Bharathi Kannan Rajendran, Jerome Ponnusamy, Vinoth Kumar Arockia Selvi, J. Department of Chemistry SRM Institute of Science and Technology Tamil Nadu Kattankulathur603203 India Kaohsiung City807 Taiwan Department of Chemistry Institute of Science Banaras Hindu University Varanasi221005 India Institute of Environmental Engineering National Sun Yat-sen University Kaohsiung804 Taiwan Center for Emerging Contaminants Research National Sun Yat-sen University Kaohsiung804 Taiwan Department of Electrical Engineering and Computer Science The University of California IrvineCA92697 United States Kaohsiung City807 Taiwan Kaohsiung Medical University Kaohsiung City807 Taiwan Department of Computational Biology Institute of Bioinformatics Saveetha School of Engineering Saveetha Institute of Medical and Technical Sciences Tamil Nadu Chennai602105 India Kaohsiung City804 Taiwan

This review discusses the most current developments and future perspectives in enzymatic and non-enzymatic glucose sensors, which have notably evolved over the preceding quadrennial period. Furthermore, a thorough exploration encompassed the sensor's intricate fabrication processes, the diverse range of materials employed, the underlying principles of detection, and an in-depth assessment of the sensors' efficacy in detecting glucose levels within essential bodily fluids such as human blood serums, urine, saliva, and interstitial fluids. It is worth noting that the accurate quantification of glucose concentrations within human blood has been effectively achieved by utilizing classical enzymatic sensors harmoniously integrated with optical and electrochemical transduction mechanisms. Monitoring glucose levels in various mediums has attracted exceptional attention from industrial to academic researchers for diabetes management, food quality control, clinical medicine, and bioprocess inspection. There has been an enormous demand for the creation of novel glucose sensors over the past ten years. research has primarily concentrated on succeeding biocompatible and enhanced sensing abilities related to the present technologies, offering innovative avenues for more effective glucose sensors. Recent developments in wearable optical and electrochemical sensors with low cost, high stability, point-of-care testing, and online tracking of glucose concentration levels in biological fluids can aid in managing and controlling diabetes globally. New nanomaterials and biomolecules that can be used in electrochemical sensor systems to identify glucose concentration levels are developed thanks to advances in nanoscience and nanotechnology. Both enzymatic and non-enzymatic glucose electrochemical sensors have garnered much interest recently and have made significant strides in detecting glucose levels. In this review, we summarise several categories of non-enzymatic glucose sensor mater

关键词： Electrocatalysts

Steering LVLMs via Sparse Autoencoder for Hallucination Mitigation

学校读者我要写书评

暂无评论

arXiv 2025年

作者： Hua, Zhenglin He, Jinghan Yao, Zijun Han, Tianxu Guo, Haiyun Jia, Yuheng Fang, Junfeng School of Computer Science and Engineering Southeast University Bangladesh Foundation Model Research Center Institute of Automation Chinese Academy of Sciences China School of Artificial Intelligence University of Chinese Academy of Sciences China Department of Computer Science and Technology Tsinghua University China Wuhan University of Technology China National University of Singapore Singapore

Large vision-language models (LVLMs) have achieved remarkable performance on multimodal tasks such as visual question answering (VQA) and image captioning. However, they still suffer from hallucinations, generating text inconsistent with visual input, posing significant risks in real-world applications. Existing approaches to address this issue focus on incorporating external knowledge bases, alignment training, or decoding strategies, all of which require substantial computational cost and time. Recent works try to explore more efficient alternatives by adjusting LVLMs’ internal representations. Although promising, these methods may cause hallucinations to be insufficiently suppressed or lead to excessive interventions that negatively affect normal semantics. In this work, we leverage sparse autoencoders (SAEs) to identify semantic directions closely associated with either hallucinations or actuality, realizing more precise and direct hallucination-related representations. Our analysis demonstrates that interventions along the faithful direction we identified can mitigate hallucinations, while those along the hallucinatory direction can exacerbate them. Building on these insights, we propose Steering LVLMs via SAE Latent Directions (SSL), a training-free method based on SAE-derived latent directions to mitigate hallucinations in LVLMs. Extensive experiments demonstrate that SSL significantly outperforms existing decoding approaches in mitigating hallucinations, while maintaining transferability across different model architectures with negligible additional time overhead. Copyright © 2025, The Authors. All rights reserved.

关键词： Visual languages

Radio-transparent dipole antenna based on a metasurface cloak

学校读者我要写书评

暂无评论

arXiv 2021年

作者： Soric, Jason Ra'di, Younes Farfan, Diego Alù, Andrea Department of Electrical and Computer Engineering University of Texas at Austin AustinTX78712 United States Photonics Initiative Advanced Science Research Center City University of New York NY10031 United States Physics Program Graduate Center City University of New York NY10016 United States Department of Electrical Engineering City College of the City University of New York NY10031 United States

Antenna technology is at the basis of ubiquitous wireless communication systems and sensors. Radiation is typically sustained by conduction currents flowing around resonant metallic objects that are optimized to enhance efficiency and bandwidth. However, resonant conductors are prone to large scattering of impinging waves, leading to challenges in crowded antenna environments due to blockage and distortion. Metasurface cloaks have been explored in the quest of addressing this challenge by reducing antenna scattering, but with limited performance in terms of bandwidth, footprint and overall scattering reduction. Here we introduce a different route towards radio-transparent antennas, in which the cloak itself acts as the radiating element, drastically reducing the overall footprint while enhancing scattering suppression and bandwidth, without sacrificing other relevant radiation metrics compared to conventional antennas. This technique offers a new application of cloaking technology, with promising features for crowded wireless communication platforms and noninvasive sensing. © 2021, CC BY-NC-ND.

关键词： Bandwidth

Metallic vs. Semiconducting properties of quasi-one-dimensional tantalum selenide van der Waals nanoribbons

学校读者我要写书评

暂无评论

arXiv 2021年

作者： Kargar, Fariborz Krayev, Andrey Wurch, Michelle Ghafouri, Yassamin Debnath, Topojit Wickramaratne, Darshana Salguero, Tina T. Lake, Roger Bartels, Ludwig Balandin, Alexander A. Center Department of Electrical and Computer Engineering University of California RiversideCA92521 United States HORIBA Scientific NovatoCA94949 United States Department of Chemistry and Material Science and Engineering Program University of California RiversideCA92521 United States Department of Chemistry University of Georgia AthensGA30602 United States Laboratory for Terahertz and Terascale Electronics Department of Electrical and Computer Engineering University of California RiversideCA92521 United States Center for Computational Materials Science U.S. Naval Research Laboratory WashingtonDC20375 United States

We conducted a tip-enhanced Raman scattering spectroscopy (TERS) and photoluminescence (PL) study of quasi-1D TaSe3-δ nanoribbons exfoliated onto gold substrates. At a selenium deficiency of δ~0.25 (Se/Ta=2.75,), the nanoribbons exhibit a strong, broad PL peak centered around ~920 nm (1.35 eV), suggesting their semiconducting behavior. Such nanoribbons revealed a strong TERS response under 785-nm laser excitation, allowing for their nanoscale spectroscopic imaging. Nanoribbons with a smaller selenium deficiency (Se/Ta=2.85, δ~0.15) did not show any PL or TERS response. The confocal Raman spectra of these samples agree with the previously-reported spectra of metallic TaSe3. The differences in the optical response of the nanoribbons examined in this study suggest that even small variations in Se content can induce changes in electronic structure, causing samples to exhibit either metallic or semiconducting character. The temperature-dependent electrical measurements of devices fabricated with both types of materials corroborate these observations. The density-functional-theory calculations revealed that incorporation of an oxygen atom in a Se vacancy can result in band gap opening and thus enable the transition from a metal to a semiconductor. However, the predicted bandgap is substantially smaller than that derived from PL data. These results indicate that the properties of van der Waals materials can vary significantly depending on stoichiometry, defect types and concentration, and possibly environmental and substrate effects. In view of this finding, local probing of nanoribbon properties with TERS becomes essential to understanding such low-dimensional systems. Copyright © 2021, The Authors. All rights reserved.

关键词： Van der Waals forces

Learning to Predict IR Drop with Effective Training for ReRAM-based Neural Network Hardware

学校读者我要写书评

暂无评论

Learning to Predict IR Drop with Effective Training for ReRA...

Design automation Conference

作者： Sugil Lee Giju Jung Mohammed E. Fouda Jongeun Lee Ahmed Eltawil Fadi Kurdahi School of Electrical and Computer Engineering UNIST Ulsan Korea Neural Processing Research Center Seoul National University Seoul Korea Department of Electrical Engineering and Computer Science University of California-Irvine CA USA

ISBN: (数字)9781728110851

ISBN: (纸本)9781728158020

Due to the inevitability of the IR drop problem in passive ReRAM crossbar arrays, finding a software solution that can predict the effect of IR drop without the need of expensive SPICE simulations, is very desirable. In this paper, two simple neural networks are proposed as software solution to predict the effect of IR drop. These networks can be easily integrated in any deep neural network framework to incorporate the IR drop problem during training. As an example, the proposed solution is integrated in BinaryNet framework and the test validation results, done through SPICE simulations, show very high improvement in performance close to the baseline performance, which demonstrates the efficacy of the proposed method. In addition, the proposed solution outperforms the prior work on challenging datasets such as CIFAR10 and SVHN.

关键词：

Assessing Dynamic Cerebral Autoregulation in Patients with Sickle Cell Disease Using Near-Infrared Spectroscopy and Paced Breathing

学校读者我要写书评

暂无评论

BLOOD 2020年 136卷 57-58页

作者： Wood, Sossena Fanta, Abeselom Dosunmu-Ogunbi, Atinuke Ruesch, Alexander Jonassaint, Jude Huppert, Theodore Novelli, Enrico M. Kainerstorfer, Jana Department of Biomedical Engineering Carnegie Mellon University Pittsburgh PA Carnegie Mellon Neuroscience Institute Carnegie Mellon University Pittsburgh PA University of Pittsburgh School of Medicine Pittsburgh PA Academic Research Center for Hematology (ARCH) University of Pittsburgh Pittsburgh PA Department of Electrical and Computer Engineering University of Pittsburgh Pittsburgh PA Heart Lung and Blood Vascular Medicine Institute University of Pittsburgh Pittsburgh PA

Introduction. Sickle cell disease (SCD) is caused by hemoglobin S polymerization leading to vaso-occlusion, small vessel disease, and reduced tissue oxygen saturation. Noninvasive assessments and screening of small vessel disease and decreased oxygen saturation in select organs, like the brain, are poorly developed. Cerebral autoregulation, a mechanism that ensures cerebral blood flow (CBF) remains constant with changes in mean arterial pressure, is impaired with and is a marker of small vessel disease. Frequency-domain near-infrared spectroscopy (FD-NIRS) can be used as a noninvasive technique to measure cerebral autoregulation by recording cerebral microvascular changes in hemoglobin concentration that are related to cerebral blood flow. Transfer function analysis is a method to noninvasively quantify the relationship between measured physiologic signals, and can be used to estimate cerebral autoregulation. In this study, we hypothesized patients with SCD have impaired cerebral autoregulation and reduced tissue oxygen saturation, which we evaluated via FD-NIRS with transfer function analysis, as compared to healthy controls. Methods. All patients provided informed consent under an IRB protocol approved by the University of Pittsburgh and Carnegie Mellon University. Patients with SCD (n = 11) and healthy, race-matched controls (n =14) were enrolled. All participants sat upright in a chair with probes of a commercial multi-distance FD-NIRS apparatus applied to their forehead in a dark room (Fig. 1A). We used paced breathing to dynamically induce controlled changes in the microvascular tone at 0.1, 0.125, and 0.1667 Hz, spanning normal to more challenging breathing rates, with a visual metronome guiding each paced breathing session (Fig.1.B). To validate each breathing pace, we recorded the participants’ respiratory rate via a belt placed around their chest, and their mean arterial pressure through a finger and arm cuff attached to a beat-to-beat finger plethysmograp

关键词：

Kernel contextual fuzzy rule model based on conditional input space partitioning driven by data reconstruction in autoencoder and randomization-based neural networks

学校读者我要写书评

暂无评论

Knowledge-Based Systems 2025年 320卷

作者： Zhang, Congcong Oh, Sung-Kwun Fu, Zunwei Pedrycz, Witold School of Mathematics and Statistics Linyi University Shandong Linyi276005 China Shandong Data Open Innovative Application Laboratory Linyi276005 China School of Automation and Electrical Engineering Linyi University Shandong Linyi276005 China School of Electrical & Electronic Engineering The University of Suwon Gyeonggi-do Hwaseong-si18323 Korea Republic of Department of Electronic Engineering Seokyeong University Seoul 02713 Korea Republic of Department of Computer Science The University of Suwon Gyeonggi-do Hwaseong-si18323 Korea Republic of Department of Electrical and Computer Engineering University of Alberta EdmontonAB T6G 2R3 Canada Systems Research Institute Polish Academy of Sciences Warsaw 00-901 Poland Istinye University Faculty of Engineering and Natural Sciences Department of Computer Engineering Sariyer/Istanbul Turkey

Fuzzy Rule-based Models (FRMs) have attracted significant interest in the field of machine learning owing to their modular architecture, robust design methodologies, and sound interpretability. This study introduces a novel Kernel Contextual Fuzzy Rule Model (KCFRM) designed to cope with regression tasks. A Kernel Contextual Fuzzy Clustering (KCFC) algorithm is proposed for conditional partitioning of the input space. Specifically, we incorporate the Mercer kernel into context fuzzy clustering to enhance the abilities of the model to distinguish, extract, and amplify useful features via nonlinear mapping, thereby generating more suitable information granules. Additionally, we employ an autoencoder's "encoding-decoding" mechanism to extract differences between data patterns and subsequently transform these into KCFC contexts via a conversion function, therefore leading to the creation of high-quality fuzzy sets. In the conclusion portion of fuzzy rules, conventional numerical or linear functions struggle to adequately describe the complex behavior present in local fuzzy regions. To mitigate this, we incorporate a Randomization-based Neural Network (RANN), capable of providing superior approximation capabilities and substantial computational efficiency. RANN overcomes traditional method constraints in representing complex behaviors within the fuzzy region. The inclusion of RANN results in more accurate and efficient conclusions for fuzzy rules. The uniqueness of this study lies in its holistic approach to designing fuzzy models with KCFC and RANN, improving the expressiveness of the rules and enhancing the generalization of the models. KCFRM's performance is assessed using various publicly available machine learning datasets, with experimental results underscoring its effectiveness and performance enhancements. © 2025

关键词： Fuzzy rules