کد مقاله | کد نشریه | سال انتشار | مقاله انگلیسی | نسخه تمام متن |
---|---|---|---|---|
7377359 | 1480111 | 2016 | 7 صفحه PDF | دانلود رایگان |
عنوان انگلیسی مقاله ISI
Minimal perceptrons for memorizing complex patterns
ترجمه فارسی عنوان
فرآیندهای مینیمم برای حفظ الگوهای پیچیده
دانلود مقاله + سفارش ترجمه
دانلود مقاله ISI انگلیسی
رایگان برای ایرانیان
کلمات کلیدی
پروپترونها، پیچیدگی شبکه، الگوهای دودویی، ذخیره سازی حافظه، معماری شبکه،
موضوعات مرتبط
مهندسی و علوم پایه
ریاضیات
فیزیک ریاضی
چکیده انگلیسی
Feedforward neural networks have been investigated to understand learning and memory, as well as applied to numerous practical problems in pattern classification. It is a rule of thumb that more complex tasks require larger networks. However, the design of optimal network architectures for specific tasks is still an unsolved fundamental problem. In this study, we consider three-layered neural networks for memorizing binary patterns. We developed a new complexity measure of binary patterns, and estimated the minimal network size for memorizing them as a function of their complexity. We formulated the minimal network size for regular, random, and complex patterns. In particular, the minimal size for complex patterns, which are neither ordered nor disordered, was predicted by measuring their Hamming distances from known ordered patterns. Our predictions agree with simulations based on the back-propagation algorithm.
ناشر
Database: Elsevier - ScienceDirect (ساینس دایرکت)
Journal: Physica A: Statistical Mechanics and its Applications - Volume 462, 15 November 2016, Pages 31-37
Journal: Physica A: Statistical Mechanics and its Applications - Volume 462, 15 November 2016, Pages 31-37
نویسندگان
Marissa Pastor, Juyong Song, Danh-Tai Hoang, Junghyo Jo,