On “Natural” Learning and Pruning in Multilayered Perceptrons
نویسندگان
چکیده
منابع مشابه
Broken symmetries in multilayered perceptrons.
The statistical mechanics of two-layered perceptrons with N input units, K hidden units, and a single output unit that makes a decision based on a majority rule (Committee Machine), is studied. Two architectures are considered. In the nonoverlapping case the hidden units do not share common inputs. In the fully connected case each hidden unit is connected to the entire input layer. In both case...
متن کاملLETTER Communicated by Sun - Ichi AmariOn \ Natural " Learning and Pruning in Multilayered PerceptronsTom
Several studies have shown that natural gradient descent for on-line learning is much more eecient than standard gradient descent. In this paper, we derive natural gradients in a slightly diierent manner and discuss implications for batch-mode learning and pruning, linking them to existing algorithms such as Levenberg-Marquardt optimization and optimal brain surgeon. The Fisher matrix plays an ...
متن کاملInitialization and optimization of multilayered perceptrons
Despite all the progress in neural networks field the technology is brittle and sometimes difficult to apply. Good initialization of adaptive parameters in neural networks and optimization of architecture are the key factor to create robust neural networks. Methods of initialization of MLPs are reviewed and new methods based on clusterization techniques are suggested. Penalty term added to the ...
متن کاملthe role of vocabulary learning strategies on vocabulary retention and on language proficiency in iranian efl students
آموزش زبان دوم طی سالهای اخیر بدنبال روشهای بهتری برای تحقق بخشیدن به اهداف معلمین و دانش آموزان بوده است . در مورد معلمین این امر منجر به تحقیقاتی در مورد ساختار زبانی، محاوره ای و تعاملی گردیده است . در مورد دانش آموزان این امر به مطالعاتی درباره نگرش دانش آموزان نسبت به فراگیری در داخل کلاس و بیرون از آن و همچنین انواع مختلف روشهای پردازش ذهنی منجر شده است . هدف از این تحقیق یافتن روشهائی اس...
15 صفحه اولذخیره در منابع من
با ذخیره ی این منبع در منابع من، دسترسی به آن را برای استفاده های بعدی آسان تر کنید
ژورنال
عنوان ژورنال: Neural Computation
سال: 2000
ISSN: 0899-7667,1530-888X
DOI: 10.1162/089976600300015637