检索结果-内蒙古大学图书馆

CSI Transactions on ICT 2025年第1期13卷 157-157页

作者： Vasanth, K. Kishore, R. K. Sugumaran, Vijayan Krishnamoorthy, Raja Ramdas, Rithvik Tadepall, Satya Kiranmai Electronics and Communication Engineering Chaitanya Bharathi Institute of Technology Hyderabad India Department of Computer Science and Engineering M. Kumarasamy College of Engineering Thavapalayam Karur India Department of Decision and Information Sciences School of Business Administration Oakland University Rochester USA Department of Electronics and Communication Engineering Kalasalingam Academy of Research and Education Krishnankoil Virudhunagar India Artificial Intelligence and Data Science Chaitanya Bharathi Institute of Technology Hyderabad India

TARGET SPECIFIC DE NOVO DESIGN OF DRUG CANDIDATE MOLECULES WITH GRAPH TRANSFORMER-BASED GENERATIVE ADVERSARIAL NETWORKS

学校读者我要写书评

暂无评论

arXiv 2023年

作者： Ünlü, Atabey Çevrim, Elif Sarıgün, Ahmet Yiğit, Melih Gökay Çelikbilek, Hayriye Bayram, Osman Güvenilir, Heval Ataş Koyaş, Altay Kahraman, Deniz Cansen Olğaç, Abdurrahman Rifaioğlu, Ahmet Banoğlu, Erden Doğan, Tunca Biological Data Science Lab Dept. of Computer Engineering Hacettepe University Turkey Dept. of Bioinformatics Graduate School of Health Sciences Hacettepe University Turkey Dept. of Chemistry Middle East Technical University Turkey Dept. of Physics Middle East Technical University Turkey Dept. of Computer Engineering Middle East Technical University Turkey Dept. Of Artificial Intelligence Engineering Bahcesehir University Turkey Cancer Systems Biology Lab Graduate School of Informatics Middle East Technical University Turkey Dept. of Pharmaceutical Chemistry Faculty of Pharmacy Gazi University Turkey Laboratory of Molecular Modeling Evias Pharmaceutical R&D Ltd Turkey Dept. of Electrical and Electronics Engineering Iskenderun Technical University Turkey Institute for Computational Biomedicine Heidelberg University Germany

Discovering novel drug candidate molecules is one of the most fundamental and critical steps in drug development. Generative deep learning models, which create synthetic data given a probability distribution, offer a high potential for designing de novo molecules. However, for them to be useful in real-life drug development pipelines, these models should be able to design drug-like and target-centric molecules. In this study, we propose an end-to-end generative system, DrugGEN, for the de novo design of drug candidate molecules that interact with intended target proteins. The proposed method represents molecules as graphs and processes them via a generative adversarial network comprising graph transformer layers. The system is trained using a large dataset of drug-like compounds and target-specific bioactive molecules to design effective inhibitory molecules against the AKT1 protein, which is critically important in developing treatments for various types of cancer. We conducted molecular docking and dynamics to assess the target-centric generation performance of the model, as well as attention score visualisation to examine model interpretability. Results indicate that our de novo molecules have a high potential for interacting with the AKT1 protein at the level of its native ligands. Using the open-access DrugGEN codebase, it is possible to easily train models for other druggable proteins, given a dataset of experimentally known bioactive molecules. © 2023, CC BY-NC-SA.

关键词： Generative adversarial networks

Integrating Appearance and Motion Cues for Visual Tracking via Spatio-Temporal Prompt

学校读者我要写书评

暂无评论

SSRN

SSRN 2025年

作者： Jia, Xianxin Hou, ZhiQiang Wang, Yunlong Yue, Hao Hu, Shuai Ma, Sugang Yang, Xiaobao Pu, Lei School of Communications and Information Engineering School of Artificial Intelligence Xi’ an University of Posts and Telecommunications Xi’an710121 China Key Laboratory of Network Data Analysis and Intelligent Processing of Shaanxi Province Xi’an710121 China School of Computer Science and Technology Xi’an University of Posts and Telecommunications Xi’an710121 China Combat Support College Rocket Force University of Engineering Xi’an710025 China

Rich appearance cues in the template frame and dynamic motion cues between consecutive frames are essential for achieving robust visual tracking. However, most existing tracking methods exhibit limitations in their architectural design, often relying on a single cue while neglecting the other. This lack of integration hinders the effective interaction between different cues, thereby restricting the model’s ability to comprehensively interpret the target characteristics. To address this challenge, we propose a unified spatio-temporal prompt learning framework, ICSPTrack, which systematically captures target appearance and motion information across consecutive ***,theAppearancecueEncoderfirstextractsappearance-relatedfeatures, followed by the Motion Cue Encoder, which derives motion-related representations. These cues are subsequently integrated within the Spatio-Temporal Prompt Decoder to facilitate future reasoning by unifying different information sources. The proposed tracker was trained in an end-to-end manner and rigorously evaluated on seven benchmark datasets, including LaSOT, TNL2K, and TrackingNet. The framework achieved success rates of 72.0%, 58.5%, and 84.4% on LaSOT, TNL2K, and TrackingNet, respectively, while attaining an average overlap rate of 74.7% on GOT-10K. Despite its low computational complexity of only 28.8 GFLOPs, the model demonstrated efficient inference at 56 FPS. Experimental results indicate that the proposed method effectively handles complex tracking scenarios and achieves excellent tracking performance. © 2025, The Authors. All rights reserved.

关键词：

Towards deployment-centric multimodal AI beyond vision and language

学校读者我要写书评

暂无评论

arXiv 2025年

作者： Liu, Xianyuan Zhang, Jiayang Zhou, Shuo van der Plas, Thijs L. Vijayaraghavan, Avish Grishina, Anastasiia Zhuang, Mengdie Schofield, Daniel Tomlinson, Christopher Wang, Yuhan Li, Ruizhe van Zeeland, Louisa Tabakhi, Sina Demeocq, Cyndie Li, Xiang Das, Arunav Timmerman, Orlando Baldwin-McDonald, Thomas Wu, Jinge Bai, Peizhen Al Sahili, Zahraa Alwazzan, Omnia Do, Thao N. Suvon, Mohammod N.I. Wang, Angeline Cipolina-Kun, Lucia Moretti, Luigi A. Farndale, Lucas Jain, Nitisha Efremova, Natalia Ge, Yan Varela, Marta Lam, Hak-Keung Celiktutan, Oya Evans, Ben R. Coca-Castro, Alejandro Wu, Honghan Abdallah, Zahraa S. Chen, Chen Danchev, Valentin Tkachenko, Nataliya Lu, Lei Zhu, Tingting Slabaugh, Gregory G. Moore, Roger K. Cheung, William K. Charlton, Peter H. Lu, Haiping Centre for Machine Intelligence University of Sheffield Sheffield United Kingdom School of Computer Science University of Sheffield Sheffield United Kingdom The Alan Turing Institute London United Kingdom Department of Metabolism Digestion and Reproduction Imperial College London London United Kingdom Department of Applied AI Simula Research Laboratory Oslo Norway Information School University of Sheffield Sheffield United Kingdom NHS England Leeds United Kingdom Institute of Health Informatics University College London London United Kingdom Department of Engineering King’s College London London United Kingdom Department of Computing Science University of Aberdeen Aberdeen United Kingdom School of Informatics University of Edinburgh Edinburgh United Kingdom School of Engineering Mathematics and Technology University of Bristol Bristol United Kingdom Department of Informatics King’s College London London United Kingdom Department of Earth Sciences University of Cambridge Cambridge United Kingdom Department of Computer Science University of Manchester Manchester United Kingdom Department of Computer Science Queen Mary University of London London United Kingdom Digital Environment Research Institute Queen Mary University of London London United Kingdom Department of Computer Science University of Bath Bath United Kingdom Department of Classics King’s College London London United Kingdom School of Electrical Electronic and Mechanical Engineering University of Bristol Bristol United Kingdom School of Engineering University of the West of England Bristol United Kingdom Cancer Research UK Scotland Institute Glasgow United Kingdom School of Business and Management Queen Mary University of London London United Kingdom City St George’s University of London London United Kingdom British Antarctic Survey Cambridge United Kingdom School of Health and Wellbeing University of Glasgow Glasgow United Kingdom Chief Data & AI Office Lloyds Banking Group Lo

Multimodal artificial intelligence (AI) integrates diverse types of data via machine learning to improve understanding, prediction, and decision-making across disciplines such as healthcare, science, and engineering. However, most multimodal AI advances focus on models for vision and language data, while their deployability remains a key challenge. We advocate a deployment-centric workflow that incorporates deployment constraints early to reduce the likelihood of undeployable solutions, complementing data-centric and model-centric approaches. We also emphasise deeper integration across multiple levels of multimodality and multidisciplinary collaboration to significantly broaden the research scope beyond vision and language. To facilitate this approach, we identify common multimodal-AI-specific challenges shared across disciplines and examine three real-world use cases: pandemic response, self-driving car design, and climate change adaptation, drawing expertise from healthcare, social science, engineering, science, sustainability, and finance. By fostering multidisciplinary dialogue and open research practices, our community can accelerate deployment-centric development for broad societal impact. © 2025, CC BY.

关键词： Decision making

Recurrent feature reasoning for image inpainting

学校读者我要写书评

暂无评论

arXiv 2020年

作者： Li, Jingyuan Wang, Ning Zhang, Lefei Du, Bo Tao, Dacheng National Engineering Research Center for Multimedia Software School of Computer Science and Institute of Artificial Intelligence Wuhan University China UBTECH Sydney AI Centre School of Computer Science Faculty of Engineering University of Sydney DarlingtonNSW2008 Australia

Existing inpainting methods have achieved promising performance for recovering regular or small image defects. However, filling in large continuous holes remains difficult due to the lack of constraints for the hole center. In this paper, we devise a Recurrent Feature Reasoning (RFR) network which is mainly constructed by a plug-and-play Recurrent Feature Reasoning module and a Knowledge Consistent Attention (KCA) module. Analogous to how humans solve puzzles (i.e., first solve the easier parts and then use the results as additional information to solve difficult parts), the RFR module recurrently infers the hole boundaries of the convolutional feature maps and then uses them as clues for further inference. The module progressively strengthens the constraints for the hole center and the results become explicit. To capture information from distant places in the feature map for RFR, we further develop KCA and incorporate it in RFR. Empirically, we first compare the proposed RFR-Net with existing backbones, demonstrating that RFR-Net is more efficient (e.g., a 4% SSIM improvement for the same model size). We then place the network in the context of the current state-of-the-art, where it exhibits improved performance. The corresponding source code is available at: https://***/jingyuanli001/ RFR-Inpainting Copyright © 2020, The Authors. All rights reserved.

关键词： Machine learning

A modified Particle Swarm Optimization algorithm with enhanced search quality and population using Hummingbird Flight patterns

学校读者我要写书评

暂无评论

Decision Analytics Journal 2023年 7卷

作者： Zare, Mohsen Akbari, Mohammad-Amin Azizipanah-Abarghooee, Rasoul Malekpour, Mostafa Mirjalili, Seyedali Abualigah, Laith Department of Electrical Engineering Faculty of engineering Jahrom University Fras Jahrom Iran Department of Electrical and Computer Engineering University of Cyprus Cyprus Green energy and mobility department RINA Tech UK Ltd Manchester United Kingdom Centre for Artificial Intelligence Research and Optimisation Torrens University Australia Brisbane 4006 QLD Australia University Research and Innovation Center Obuda University Budapest 1034 Hungary Computer Science Department Prince Hussein Bin Abdullah College for Information Technology Al AlBayt University Mafraq 130040 Jordan Hourani Center for Applied Scientific Research Al-Ahliyya Amman University Amman 19328 Jordan MEU Research Unit Middle East University Amman 11831 Jordan Applied science research center Applied science private university Amman 11931 Jordan School of Computer Sciences Universiti Sains Malaysia Pulau Pinang 11800 Malaysia School of Engineering and Technology Sunway University Malaysia Petaling Jaya 27500 Malaysia

This study proposes a modified Particle Swarm Optimization (PSO) algorithm based on Hummingbird Flight (HBF) patterns to enhance the search quality and population diversity. The HBF has five concepts: (1) Smaller steps toward position updating are more likely than larger ones, (2) Position changes are made step by step throughout the flight, (3) Flight energy is conserved during the nectar-searching process, (4) Hummingbirds do not fly in large groups in confined spaces, and (5) Simultaneous position changes in all directions are not realistic. A comprehensive study on two CEC-2010 and CEC-2013 benchmark suites is conducted to verify the effectiveness of the proposed PSO-HBF algorithm. The proposed algorithm is also evaluated and compared to other well-known PSO algorithms using shifted and rotated CEC 2005 and CEC 2014 benchmark functions. Four cases in economic dispatch, the 10-unit reserve constraint, and the 30-unit dynamic economic dispatch (DED) are further examined. The last two cases investigate how the proposed PSO-HBF deals with large-scale practical problems. The results demonstrated that the PSO-HBF algorithm is superior to seven other modified algorithms, improving eight and ten functions on the 2010 and 2013 benchmarks, respectively. Furthermore, achieving the third rank among the nineteen improved PSO algorithms based on the 2005 functions confirms the effectiveness of the proposed algorithm. Moreover, in two cases of the DED problem, the results of PSO-HBF show significant improvement over previously published papers. The PSO-HBF algorithm's source code can be accessed publicly at http://***/projects/psohbf. © 2023 The Author(s)

关键词： Algorithm Becnhmark Dynamic economic dispatch Economic dispatch Evolutionary algorithm Hummingbird flight Large-scale optimization Optimization Particle swarm optimization

Semantics-Guided Contrastive Network for Zero-Shot Object Detection

学校读者我要写书评

暂无评论

arXiv 2021年

作者： Yan, Caixia Chang, Xiaojun Luo, Minnan Liu, Huan Zhang, Xiaoqin Zheng, Qinghua The Ministry of Education Key Lab of Intelligent Networks and Network Security National Engineering Lab for Big Data Analytics School of Electronic and Information Engineering Xi'an Jiaotong University Shaanxi710049 China The ReLER Lab AAAI Faculty of Engineering and Information Technology University of Technology Sydney Australia The College of Computer Science and Artificial Intelligence Wenzhou University China

Zero-shot object detection (ZSD), the task that extends conventional detection models to detecting objects from unseen categories, has emerged as a new challenge in computer vision. Most existing approaches tackle the ZSD task with a strict mapping-transfer strategy that may lead to suboptimal ZSD results: 1) the learning process of these models neglects the available semantic information on unseen classes, which can easily bias towards the seen categories;2) the original visual feature space is not well-structured for the ZSD task due to the lack of discriminative information. To address these issues, we develop a novel Semantics-Guided Contrastive Network for ZSD, named ContrastZSD, a detection framework that first brings contrastive learning mechanism into the realm of zero-shot detection. Particularly, ContrastZSD incorporates two semantics-guided contrastive learning subnets that contrast between region-category and region-region pairs respectively. The pairwise contrastive tasks take advantage of supervision signals derived from both the ground truth label and class similarity information. By performing supervised contrastive learning over those explicit semantic supervision, the model can learn more knowledge about unseen categories to avoid the bias problem to seen concepts, while optimizing the visual data structure to be more discriminative for better visual-semantic alignment. Extensive experiments are conducted on two popular benchmarks for ZSD, i.e., PASCAL VOC and MS COCO. Results show that our method outperforms the previous state-of-the-art on both ZSD and generalized ZSD tasks. © 2021, CC BY.

关键词： Object detection

Hybrid feature selection to identify important single nucleotide polymorphism for asthma classification

学校读者我要写书评

暂无评论

AIP Conference Proceedings 2024年第1期2991卷

作者： Wan Ni Wong Weng Howe Chan Hui Wen Nies Kauthar Mohd Daud Artificial Intelligence and Bioinformatics Group School of Computing Faculty of Engineering Universiti Teknologi Malaysia 81310 Skudai Johor Malaysia UTM Big Data Centre Ibnu Sina Institute for Scientific and Industrial Research Universiti Teknologi Malaysia 81310 Skudai Johor Malaysia Center for Artificial Intelligence Technology Faculty of Information Science & Technology Universiti Kebangsaan Malaysia 43600 UKM Bangi Selangor Malaysia

Single nucleotide polymorphism (SNP), being the most common genetic variation, is found to be responsible for traits and complex diseases such as asthma. Polygenic risk scoring (PRS) and machine learning (ML) are the current two common types of genetic risk prediction models. PRS employs statistical association testing to identify alleles with risk. However, it was found not performing well in unraveling the SNP-SNP interactions and attaining low predictive value. In contrast, ML algorithms are more robust in detecting disease associated SNPs, resulting in an improvement in prediction performance. The challenge remains as the high dimensionality nature of SNP tends to cause overfitting problems and impact the classifier performance. This paper introduces a two-stage hybrid feature selection that integrates Mutual Information and Random Forest Recursive Feature Elimination for identifying informative SNPs from 176,287 SNPs in a dataset of 128 asthmatic and non-asthmatic samples. The reduced samples based on the identified informative SNPs are then used with Support Vector Machine to train a classification model. The model from this paper outperforms other ML models with an average accuracy, precision, recall and area under the curve of 98.91%, 98.36%, 99.72% and 0.99 respectively. Additional analysis shows that the genes associated with the top selected SNPs, namely Ring Finger Protein 217 (RNF217) and A-kinase-anchoring protein 6 (AKAP6) are correlated with respiratory diseases. This shows that the proposed method can identify important SNPs that are able to differentiate between asthmatic and non-asthmatic samples.

关键词：

Pay-Per-Flight Dynamic Pricing of UAV Operations

学校读者我要写书评

暂无评论

Pay-Per-Flight Dynamic Pricing of UAV Operations

International Conference on artificial intelligence and data Analytics for Air Transportation (AIDA-AT)

作者： Jaime Rubio-Hervas Abhishek Gupta Yew-Soon Ong Mahmut Reyhanoglu Data Science and Artificial Intelligence Research Centre Nanyang Technological University Singapore Singapore Institute of Manufacturing Technology Agency for Science Technology and Research Singapore School of Computer Science and Engineering Nanyang Technological University Singapore University of North Carolina at Asheville Asheville USA

ISBN: (数字)9781728153803

ISBN: (纸本)9781728153810

Insuring unmanned aerial vehicles (UAVs) is a relatively new concept, where not much data is available yet. We propose the combination of available data from different sources, other than past accident rates, to stochastically model the operational environment by using Gaussian process-based function approximations. A data-driven risk measure is then derived through such stochastic formulation accounting for both aleatoric uncertainties of the considered environmental factors as well as epistemic uncertainties originating from the geographical sparsity of data collection sources. The risk measure is obtained in a path-integral form which represents the operational risk associated with a defined operation in partially unknown environments. A novel pay-per-flight dynamic pricing scheme is derived from such risk measure.

关键词： Hazards Accidents Pricing Stochastic processes Measurement Insurance Atmospheric modeling

Contract-based time-of-use pricing for energy storage investment

学校读者我要写书评

暂无评论

arXiv 2020年

作者： Zhao, Dongwei Wang, Hao Huang, Jianwei Lin, Xiaojun Department of Information Engineering Chinese University of Hong Kong Hong Kong Department of Data Science and Artificial Intelligence Monash University Australia School of Science and Engineering Chinese University of Hong Kong Shenzhen Institute of Artificial Intelligence and Robotics for Society Shenzhen Hong Kong School of Electrical and Computer Engineering Purdue University United States

Time-of-use (ToU) pricing is widely used by the electricity utility. A carefully designed ToU pricing can incentivize end-users’ energy storage deployment, which helps shave the system peak load and reduce the system social cost. However, the optimization of ToU pricing is highly non-trivial, and an improperly designed ToU pricing may lead to storage investments that are far from the social optimum. In this paper, we aim at designing the optimal ToU pricing, jointly considering the social cost of the utility and the storage investment decisions of users. Since the storage investment costs are users’ private information, we design low-complexity contracts to elicit the necessary information and induce the proper behavior of users’ storage investment. The proposed contracts only specify three contract items, which guides users of arbitrarily many different storage-cost types to invest in full, partial, or no storage capacity with respect to their peak demands. Our contracts can achieve the social optimum when the utility knows the aggregate demand of each storage-cost type (but not the individual user’s type). When the utility only knows the distribution of each storage-cost type’s demand, our contracts can lead to a near-optimal solution. The gap with the social optimum is as small as 1.5% based on the simulations using realistic data. We also show that the proposed contracts can reduce the system social cost by over 30%, compared with no storage investment benchmark. Copyright © 2020, The Authors. All rights reserved.

关键词： Investments