检索结果-内蒙古大学图书馆

An Identification of Guava Fruit Disease Using ML

学校读者我要写书评

暂无评论

An Identification of Guava Fruit Disease Using ML

International Conference on Advancement in Electrical and Electronic engineering (ICAEEE)

作者： Khandakar Rabbi Ahmed Towfika Salam Subarna Nandi Nazmul Hasan Omor Faruk Al Rashid Shafiun Miraz Md. Shafi Reza Farjana Rahman Miyan Research Institute International University of Business Agriculture and Technology Uttara Model Town Dhaka Bangladesh Dept. of Mechatronics Engineering Rajshahi University of Engineering and Technology (RUET) Kazla Rajshahi Bangladesh Dept. of Computer Science and Engineering Rajshahi University of Engineering and Technology (RUET) Rajshahi Bangladesh Dept. of Data Science London South Bank University 103 Borough Rd London Dept. of Computer Science and Engineering Varendra University Rajshahi Bangladesh Dept. of Computer Science and Engineering International University of Business Agriculture and Technology Dhaka Bangladesh Graduate School of Integrated Energy of AI Jeonbuk National University Jeonju North Jeolla South Korea

ISBN: (数字)9798350388282

ISBN: (纸本)9798350388299

Fruits contain a lot of fiber. Fruits contain biologically active substances that improve our health. This study focuses on identifying and classifying guava fruit diseases. Guava disease has become a significant issue for guava production, threatening farmers’ socioeconomic growth. As a result of this phenomenon, an automated computer vision-based guava disease diagnosis system was developed that can detect malicious guava and direct it to early cure approaches, reducing relative economic loss. In light of this, we present in this study a guava illness detection and categorization recommendation approach based on convolutional neural networks (CNN). This paper employs transfer learning to classify images using our customized convolutional neural network (CNN) model architecture. The photographs in the data set were obtained from the Mendeley data website. The data set includes 306 images of healthy and unhealthy guava fruits and leaves. The four disease classes targeted by the data sets are dot, Canker, mummification, and Rust. We obtained an average accuracy of 74.72% for training and 34.74% for testing from the data set used.

关键词： Training Accuracy Transfer learning Production Computer architecture Optical fiber networks Convolutional neural networks

How Multimodal Integration Boost the Performance of LLM for Optimization: Case Study on Capacitated Vehicle Routing Problems

学校读者我要写书评

暂无评论

How Multimodal Integration Boost the Performance of LLM for ...

Multidisciplinary computational Intelligence Incubators (MCII), IEEE Symposium for

作者： Yuxiao Huang Wenjie Zhang Liang Feng Xingyu Wu Kay Chen Tan Dept. of Data Science and Artificial Intelligence (DSAI) The Hong Kong Polytechnic University Hong Kong SAR Dept. of Electrical and Elctronics Engineering The Hong Kong Polytechnic University Hong Kong SAR College of Computer Science Chongqing University Chongqing China Dept. of DSAI The Hong Kong Polytechnic University Hong Kong SAR

ISBN: (数字)9798331508395

ISBN: (纸本)9798331508401

Recently, large language models (LLMs) have notably positioned them as capable tools for addressing complex optimization challenges. Despite this recognition, a predominant limitation of existing LLM-based optimization methods is their struggle to capture the relationships among decision variables when relying exclusively on numerical text prompts, especially in high-dimensional problems. Keeping this in mind, we first propose to enhance the optimization performance using multimodal LLM capable of processing both textual and visual prompts for deeper insights of the processed optimization problem. This integration allows for a more comprehensive understanding of optimization problems, akin to human cognitive processes. We have developed a multimodal LLM-based optimization framework that simulates human problem-solving workflows, thereby offering a more nuanced and effective analysis. The efficacy of this method is evaluated through extensive empirical studies focused on a well-known combinatorial optimization problem, i.e., capacitated vehicle routing problem. The results are compared against those obtained from the LLM-based optimization algorithms that rely solely on textual prompts, demonstrating the significant advantages of our multimodal approach.

关键词： Visualization Cognitive processes Large language models Vehicle routing Optimization methods Multisensory integration Problem-solving Optimization computational intelligence

Drug Recommendation System Using Patient Reviews Based on Sentimental Analysis 2

学校读者我要写书评

暂无评论

Drug Recommendation System Using Patient Reviews Based on Se...

2nd International Conference on Machine Learning and Autonomous Systems, ICMLAS 2025

作者： Kumar, C Siva Sagar, P.Lakshmi Kumar, Golla Dinesh Reddy, Chereddy Harshitha Krushna, Bandlamudi Sai Reddy, Chikkem Mahidhar Sree Vidyanikethan Engineering College Dept. of CSSE AP Tirupati517102 India Department of Data Science A.P Tirupati India Sree Vidyanikethan Engineering College AP Tirupati517102 India

ISBN: (纸本)9798331505745

The integration of digital technology in healthcare has been greatly facilitated by the internet. It has made access to vital data quite easy, with which solutions created help in improving patient outcomes. This leads to an improvement in patient care and medical procedures. Vigilance is required to avoid the dissemination of information that may be incorrect or dangerous, and thus harm patients and healthcare providers alike. The amount of patient-generated data involved is huge, and this has become quite central to the drug recommendation system in the healthcare industry. It is envisioned that medication recommendations by these systems are highly precise to increase efficacy and accuracy of treatment. Using Natural Language Processing tools, the suggested approach will make a prediction about the best drug, using ratings and reviews made by the patients. It would be advisable to apply some NLP approaches to the input provided by the patient to extract some useful insights and information that might help in prescription decisions. Examples of some machine learning methods applied for enhancing accuracy and dependability of such drug recommendations include the XGBoost classifier and LightGBM model. These algorithms further process and analyse data to bring out the relationships and patterns not otherwise evident. They therefore provide crucial help toward the formation and establishment of an effective or reliable drug-related recommendation system and greatly improved the treatment of patients. © 2025 IEEE.

关键词： Drug infusion

Robust Expected Information Gain for Optimal Bayesian Experimental Design Using Ambiguity Sets

学校读者我要写书评

暂无评论

arXiv 2022年

作者： Go, Jinwoo Isaac, Tobin Computational Science and Engineering Dept. Georgia Institute of Technology AtlantaGA United States

The ranking of experiments by expected information gain (EIG) in Bayesian experimental design is sensitive to changes in the model’s prior distribution, and the approximation of EIG yielded by sampling will have errors similar to the use of a perturbed prior. We define and analyze robust expected information gain (REIG), a modification of the objective in EIG maximization by minimizing an affine relaxation of EIG over an ambiguity set of distributions that are close to the original prior in KL-divergence. We show that, when combined with a sampling-based approach to estimating EIG, REIG corresponds to a ‘log-sum-exp’ stabilization of the samples used to estimate EIG, meaning that it can be efficiently implemented in practice. Numerical tests combining REIG with variational nested Monte Carlo (VNMC), adaptive contrastive estimation (ACE) and mutual information neural estimation (MINE) suggest that in practice REIG also compensates for the variability of under-sampled estimators. © 2022, CC BY.

关键词： Design of experiments

QuAKE: Speeding up Model Inference Using Quick and Approximate Kernels for Exponential Non-Linearities

学校读者我要写书评

暂无评论

arXiv 2024年

作者： Narayanaswami, Sai Kiran Srinivasan, Gopalakrishnan Ravindran, Balaraman Centre for Responsible AI Indian Institute of Technology Madras India Robert Bosch Centre for Data Science and AI Dept. of Computer Science and Engineering Indian Institute of Technology Madras India Wadhwani School of Data Science and AI Indian Institute of Technology Madras India

As machine learning gets deployed more and more widely, and model sizes continue to grow, improving computational efficiency during model inference has become a key challenge. In many commonly used model architectures, including Transformers, a significant portion of the inference computation is comprised of exponential non-linearities such as Softmax. In this work, we develop QuAKE, a collection of novel operators that leverage certain properties of IEEE-754 floating point representations to quickly approximate the exponential function without requiring specialized hardware, extra memory, or precomputation. We propose optimizations that enhance the efficiency of QuAKE in commonly used exponential non-linearities such as Softmax, GELU, and the Logistic function. Our benchmarks demonstrate substantial inference speed improvements between 10% and 35% on server CPUs, and 5% and 45% on embedded and mobile-scale CPUs for a variety of model architectures and sizes. Evaluations of model performance on standard datasets and tasks from various domains show that QuAKE operators are able to provide sizable speed benefits with little to no loss of performance on downstream tasks. Copyright © 2024, The Authors. All rights reserved.

关键词： Exponential functions

A Study on Neuro-Symbolic Artificial Intelligence: Healthcare Perspectives

学校读者我要写书评

暂无评论

arXiv 2025年

作者： Hossain, Delower Chen, Jake Y. Dept. of Computer Science The University of Alabama Birmingham United States Dept. of Genetic Biomedical Engineering Informatics and Data Science Heersink School of Medicine The University of Alabama Birmingham United States Systems Pharmacology AI Research Center The University of Alabama Birmingham United States

Over the last few decades, Artificial Intelligence (AI) scientists have been conducting investigations to attain human-level performance by a machine in accomplishing a cognitive task. Within machine learning, the ultimate aspiration is to attain Artificial General Intelligence (AGI) through a machine. This pursuit has led to the exploration of two distinct AI paradigms. Symbolic AI, also known as classical or GOFAI (Good Old-Fashioned AI) and Connectionist (Sub-symbolic) AI, represented by Neural Systems, are two mutually exclusive paradigms. Symbolic AI excels in reasoning, explainability, and knowledge representation but faces challenges in processing complex real-world data with noise. Conversely, deep learning (Black-Box systems) research breakthroughs in neural networks are notable, yet they lack reasoning and interpretability. Neuro-symbolic AI (NeSy), an emerging area of AI research, attempts to bridge this gap by integrating logical reasoning into neural networks, enabling them to learn and reason with symbolic representations. While a long path, this strategy has made significant progress towards achieving common sense reasoning by systems. This article conducts an extensive review of over 977 studies from prominent scientific databases (DBLP, ACL, IEEExplore, Scopus, PubMed, ICML, ICLR), thoroughly examining the multifaceted capabilities of Neuro-Symbolic AI, with a particular focus on its healthcare applications, particularly in drug discovery, and Protein engineering research. The survey addresses vital themes, including reasoning, explainability, integration strategies, 41 healthcare-related use cases, benchmarking, datasets, current approach limitations from both healthcare and broader perspectives, and proposed novel approaches for future experiments. Moreover, it identifies critical challenges in applying Neuro-Symbolic AI to medicine and represents a comprehensive exploration of its transformative potential in the biomedical field. © 2025, CC BY-NC

关键词： Knowledge representation

Malicious Traffic Classification in WSN using Deep Learning Approaches 2

学校读者我要写书评

暂无评论

Malicious Traffic Classification in WSN using Deep Learning ...

2nd IEEE International Conference on Communication, Security and Artificial Intelligence, ICCSAI 2023

作者： Gupta, Chhote Lal Prasad Rajassekharan, Dinesh Sharma, Dilip Kumar Elangovan, Mohanraj Myilsamy, Varatharaj Upreti, Kamal Bansal Institute of Engineering and Technology Dept. of Computer Science and Engineering Lucknow India Peninsula College Faculty of Computer Science Peninsula Malaysia Jaypee University of Engineering and Technology Department of Mathematics Madhya Pradesh Guna473226 India K. S. Rangasamy College of Technology Department of Artificial Intelligence and Data Science Tamil Nadu Tiruchengode637215 India V.S.B College of Engineering Technical Campus Department of Electrical and Electronics Engineering Tamil Nadu Coimbatore642109 India Department of Computer Science Delhi-NCR Ghaziabad India

ISBN: (纸本)9798350369960

Classifying malicious traffic in Wireless Sensor Networks (WSNs) is crucial for maintaining the network's security and dependability. Traditional security techniques are challenging to deploy in WSNs because they comprise tiny, resourceconstrained components with limited processing and energy capabilities. On the other hand, machine learning-based techniques, such as Deep Learning (DL) models like LSTMs, may be used to detect and categorize fraudulent traffic accurately. The classification of malicious traffic in WSNs is crucial because of security. To protect the network's integrity, data, and performance and ensure the system functions properly and securely for its intended use, hostile traffic categorization in WSNs is essential. Classifying malicious communication in a WSN using a Long Short-Term Memory (LSTM) is efficient. WSNs are susceptible to several security risks, such as malicious nodes or traffic that can impair network performance or endanger data integrity. In sequential data processing, LSTM is a Recurrent Neural Network (RNN) appropriate for identifying patterns in network traffic data. © 2023 IEEE.

关键词： data routing deep learning LSTM malicious transmission neural network Traffic data transmission WSN

Temporal Planning via Interval Logic Satisfiability for Autonomous Systems

学校读者我要写书评

暂无评论

arXiv 2024年

作者： Ramirez, Miquel Singh, Anubhav Stuckey, Peter J. Manzie, Chris The University of Melbourne Dept. of Electrical and Electronic Engineering ParkvilleVIC3052 Australia The University of Toronto Dept. of Mechanical and Industrial Engineering TorontoONM5S 3G8 Canada Monash University Dept. of Data Science & AI ClaytonVIC3168 Australia

Many automated planning methods and formulations rely on suitably designed abstractions or simplifications of the constrained dynamics associated with agents to attain computational scalability. We consider formulations of temporal planning where intervals are associated with both action and fluent atoms, and relations between these are given as sentences in Allen's Interval Logic. We propose a notion of planning graphs that can account for complex concurrency relations between actions and fluents as a Constraint Programming (CP) model. We test an implementation of our algorithm on a state-of-the-art framework for CP and compare it with PDDL 2.1 planners that capture plans requiring complex concurrent interactions between agents. We demonstrate our algorithm outperforms existing PDDL 2.1 planners in the case studies. Still, scalability remains challenging when plans must comply with intricate concurrent interactions and the sequencing of actions. © 2024, CC BY-NC-SA.

关键词： Scalability

FLODA: Harnessing Vision-Language Models for Deepfake Assessment

学校读者我要写书评

暂无评论

FLODA: Harnessing Vision-Language Models for Deepfake Assess...

IEEE International Conference on Consumer Electronics (ICCE)

作者： Seunghyeon Park Youngho Bae Gunhui Han Alexander W Olson School of Integrated Technology Yonsei University Incheon Republic of Korea dept. Data Science Hanyang University Seoul Republic of Korea dept. Mechanical Engineering Yonsei University Seoul Republic of Korea Centre for Analytics and Artificial Intelligence Engineering University of Toronto Toronto Canada

ISBN: (数字)9798331521165

ISBN: (纸本)9798331521172

The proliferation of advanced generative AI models has ushered in a new era of image creation, where the distinction between real and synthetic content has become increasingly blurred. This phenomenon poses significant challenges to the integrity and trustworthiness of digital content and AI systems. Recently, Vision-Language Models (VLMs) have shown great potential in addressing these challenges due to their superior performance in vision-language tasks. In this paper, we propose FLODA (FLorence-2 Optimized for Deepfake Assessment), an advanced method designed to surpass existing deepfake detection models. FLODA leverages the VLM model, Florence-2, and extends it by integrating image captioning and authenticity assessment into a single end-to-end architecture. By utilizing caption information, FLODA enhances visual analysis with contextual details, streamlining both caption generation and deepfake detection. To ensure FLODA's efficacy, we performed an ablation study to identify the optimal model configuration and demonstrated its contributions to performance improvements. We also compared our best FLODA model against existing benchmarks through rigorous evaluation. Notably, FLODA demonstrates strong generalization, showcasing its robustness across diverse scenarios with an average accuracy of 97.14%. This research advances digital content integrity and AI trustworthiness, providing a promising direction for future VLM applications. Code and models can be found at https://***/byh711/FLODA.

关键词： Training Deepfakes Visualization Accuracy Generative AI Streaming media Robustness Faces Resilience Context modeling