0% Complete
English
صفحه اصلی
/
شانزدهمین کنفرانس بین المللی فناوری اطلاعات و دانش
Embedding-Consistent Contrastive Learning: A Robust Approach for Imbalanced Classification
نویسندگان :
Sobhan Siamak
1
Eghbal Mansoori
2
1- دانشگاه شیراز
2- دانشگاه شیراز
کلمات کلیدی :
contrastive learning،deep learning،imbalanced classification،representation learning،medical image analysis
چکیده :
This research introduces Embedding-Consistent Contrastive Learning (EC-CL), a novel framework that leverages contrastive learning to alleviate the adverse effects of class imbalance in medical image analysis. EC-CL architecturally enforces a structured latent geometry, moving beyond simple instance discrimination to ensure both intra-class compactness and inter-class separation within the embedding manifold through a specialized cosine-based objective function. The framework's performance was analytically evaluated on a comprehensive suite of six medical imaging sub-datasets from the MedMNIST benchmark—including BloodMNIST, DermaMNIST, OrganCMNIST, OrganSMNIST, RetinaMNIST, and PneumoniaMNIST. Experimental results demonstrate that EC-CL consistently outperformed state-of-the-art counterparts, achieving an average improvement of 4.3% in AUC and 3.8% in accuracy on multi-class tasks. A notable advantage is that it secured these superior results using lower-resolution inputs, underscoring its computational efficiency. This work establishes EC-CL as an exceptional and principled paradigm for imbalanced classification, effectively bridging robust representation learning with the demands of real-world clinical deep learning applications.
لیست مقالات
لیست مقالات بایگانی شده
تاثیر مدیریت دانش مشتری بر توسعه محصول جدید و نوآورانه با رویکرد مدل سازی معادلات ساختاری با استفاده از حداقل مربعات جزئی: مطالعۀ موردی شرکت کاله
دکتر آرش خسروی - سیده فاطمه حسینی - دکتر مرتضی رجب زاده آرش خسروی - سیده فاطمه حسینی - مرتضی رجب زاده -
A Multi Objective & Trust-Based Workflow Scheduling Method In Cloud Computing Based On The MVO Algorithm
Fatemeh Ebadifard
Face Recognition Based on Local Statistical Features and Artificial Neural Network
Mehdi Moghimi - Dr Hadi Grailu
Design and Simulation of a New Multiplexer with Energy Analysis in Quantum Cellular Automata Technology
- - -
A Joint Trajectory and Energy Harvesting Method for an UAV Enabled Disaster Response Network
Hosein Mohammadi Firozjae - Javad Zeraatkar Moghaddam - Mehrdad Ardebilipour
SecVanet: provably secure authentication protocol for sending emergency events in VANET
Seyed Amir Mousavi - Mohammad Sadeq Sirjani - Seyyed Javad Bozorg zadeh Razavi - Morteza Nikooghadam
A Deep Learning Framework for Phase-Aware Feature Representation to Improve Sound Source Direction and Distance Estimation
Zahra Abolfazli - Hamid Reza Abutalebi
Stock Market Prediction Using Hard and Soft Data Fusion
Saeed Mohammadi Dashtaki - Masoud Alizadeh - Behzad Moshiri
مروری بر الگوریتمهای انتخاب مشتری در یادگیری فدرال
عطیه منعمی بیدگلی - رضا مهدوی
Intelligent Transportation System (ITS) Using Internet of Things (IoT)
Engineer Reza Khalilian - Dr. Abdalhossein Rezai - Dr. Sayyed Mohammad Reza Talakesh
بیشتر
ثمین همایش، سامانه مدیریت کنفرانس ها و جشنواره ها - نگارش 42.5.2