Evolution Strategies: An Alternative to Gradient Based Learning
نویسنده
چکیده
Learning in feedforward neural networks may fail due to several reasons. We give a model for the failure caused by premature saturation of hidden neurons. In order to avoid this type of failure, we suggest evolution strategies for the training of the networks. The properties of the chosen algorithm are demonstrated by some applications.
منابع مشابه
Efficient Natural Evolution Strategies Evolution Strategies and Evolutionary Programming Track
Efficient Natural Evolution Strategies (eNES) is a novel alternative to conventional evolutionary algorithms, using the natural gradient to adapt the mutation distribution. Unlike previous methods based on natural gradients, eNES uses a fast algorithm to calculate the inverse of the exact Fisher information matrix, thus increasing both robustness and performance of its evolution gradient estima...
متن کاملNationalization of an educational reform: evolution of training and development in research
Many medical universitiesaround the world have donestunning changes in the recent years.Some of these changes have been performed with the support of international organizations.They changed the traditionaltrainingprograms based on teacher- centered learning and disciplined approach to the educational course based on community needs responsivity and Student-centered learning. This process is ai...
متن کاملDeep Neuroevolution: Genetic Algorithms Are a Competitive Alternative for Training Deep Neural Networks for Reinforcement Learning
Deep artificial neural networks (DNNs) are typically trained via gradient-based learning algorithms, namely backpropagation. Evolution strategies (ES) can rival backprop-based algorithms such as Q-learning and policy gradients on challenging deep reinforcement learning (RL) problems. However, ES can be considered a gradient-based algorithm because it performs stochastic gradient descent via an ...
متن کاملMulti-objective Differential Evolution for the Flow shop Scheduling Problem with a Modified Learning Effect
This paper proposes an effective multi-objective differential evolution algorithm (MDES) to solve a permutation flow shop scheduling problem (PFSSP) with modified Dejong's learning effect. The proposed algorithm combines the basic differential evolution (DE) with local search and borrows the selection operator from NSGA-II to improve the general performance. First the problem is encoded with a...
متن کاملBack to Basics: Benchmarking Canonical Evolution Strategies for Playing Atari
Evolution Strategies (ES) have recently been demonstrated to be a viable alternative to reinforcement learning (RL) algorithms on a set of challenging deep RL problems, including Atari games and MuJoCo humanoid locomotion benchmarks. While the ES algorithms in that work belonged to the specialized class of natural evolution strategies (which resemble approximate gradient RL algorithms, such as ...
متن کاملذخیره در منابع من
با ذخیره ی این منبع در منابع من، دسترسی به آن را برای استفاده های بعدی آسان تر کنید
عنوان ژورنال:
دوره شماره
صفحات -
تاریخ انتشار 2007