using game theory techniques in self-organizing maps training
thesis
- وزارت علوم، تحقیقات و فناوری - دانشکده علوم اقتصادی - دانشکده مدیریت
- author احسان شکاری
- adviser محمد باقر منهاج حسن رنگریز
- publication year 1393
abstract
شبکه خود سازمانده پرکاربردترین شبکه عصبی برای انجام خوشه بندی و کوانتیزه نمودن برداری است. از زمان معرفی این شبکه تاکنون، از این روش در مسائل مختلف در حوزه های گوناگون استفاده و توسعه ها و بهبودهای متعددی برای آن ارائه شده است. شبکه خودسازمانده از تعدادی سلول برای تخمین تابع توزیع الگوهای ورودی در فضای چندبعدی استفاده می کند. احتمال وجود سلول مرده مشکلی اساسی در الگوریتم شبکه خودسازمانده به حساب می آید. مقداردهی اولیه نامناسب به بردارهای وزن سلول ها و محدب نبودن شکل توزیع ورودی دلایل اصلی به وجود آمدن سلول های مرده هستند. در این پایان نامه، از مفاهیم نظریه بازی استفاده و یک شبکه جدید خودسازمانده مبتنی بر نظریه بازی، به منظور بهبود تخمین تابع توزیع ورودی و حل مشکل سلول مرده، ارائه شده است. هر سلول به عنوان یک بازیکن با یک مجموعه استراتژی در نظر گرفته می شود. در طول آموزش، بازیکنان بر سر تصاحب تعداد بیشتری از الگوهای ورودی باهم رقابت می کنند. اجرای روش پیشنهادی بر روی مجموعه داده های مختلف، کارا بودن این روش را نشان می دهد.
similar resources
Green Product Consumers Segmentation Using Self-Organizing Maps in Iran
This study aims to segment the market based on demographical, psychological, and behavioral variables, and seeks to investigate their relationship with green consumer behavior. In this research, self-organizing maps are used to segment and to determine the features of green consumer behavior. This was a survey type of research study in which eight variables were selected from the demographical,...
full textGTSOM: Game Theoretic Self-organizing Maps
Self-Organizing Maps (SOM) is a powerful tool for clustering and discovering patterns in data. Input vectors are compared to neuron weight vectors to form the SOM structure. An update of a neuron only benefits part of the feature map, which can be thought of as a local optimization problem. A global optimization model could improve representation to data by a SOM. Game Theory is adopted to anal...
full textVisualization techniques of self-organizing maps
Neural networks try, in a computing way, to simulate human brain, including its behavior, by making errors and learning and thereby making new discovers. Self-organizing maps are part of a neural network group based on competitive networks where competition is used as a way of learning. They try to find similarities between data, based only on input data, grouping similar data to each other and...
full textprediction of stock market crash using self-organizing maps
market crash is a phenomenon which occurs in stock markets occasionally and leads to loss of the investors’ wealth and assets in a relatively short period of time. therefore, attempts for prediction of this phenomenon are of much importance for the investors, financial institutions and government. to this date, numerous and varied studies have been carried out for predicting and modeling stock...
full textInfer Species Phylogenies Using Self-Organizing Maps
With rapid advances in genomics, phylogenetics has turned to phylogenomics due to the availability of large amounts of sequence and genome data. However, incongruence between species trees and gene trees remains a challenge in molecular phylogenetics for its biological and algorithmic complexities. A state-of-the-art gene concatenation approach was proposed to resolve this problem by inferring ...
full textInteractive Face Retrieval using Self-Organizing Maps
An interactive face retrieval system that uses selforganizing maps and user feedback is described. The system solves some problems of related content-based image retrieval systems: non-existence of trivial high-level human descriptions of the images and the gap between the high-level descriptions and the low-level features used to index the images.
full textMy Resources
document type: thesis
وزارت علوم، تحقیقات و فناوری - دانشکده علوم اقتصادی - دانشکده مدیریت
Keywords
Hosted on Doprax cloud platform doprax.com
copyright © 2015-2023