Evolving Neural Turing Machines
نویسندگان
چکیده
Instead of training a Neural Turing Machine (NTM) with gradient descent [1], in this work NTMs are trained through an evolutionary algorithm. Preliminary results suggest that this setup can greatly simplify the neural model, generalizes better, and does not require accessing the entire memory content at each time-step. We show preliminary results on a simple copy and T-Maze learning task.
منابع مشابه
Some improvements in fuzzy turing machines
In this paper, we improve some previous definitions of fuzzy-type Turing machines to obtain degrees of accepting and rejecting in a computational manner. We apply a BFS-based search method and some level’s upper bounds to propose a computational process in calculating degrees of accepting and rejecting. Next, we introduce the class of Extended Fuzzy Turing Machines equipped with indeterminacy s...
متن کاملInteractive Evolving Recurrent Neural Networks Are Super-Turing Universal
We consider a model of evolving recurrent neural networks where the synaptic weights can change over time, and we study the computational power of such networks in a basic context of interactive computation. In this framework, we prove that both models of rationaland real-weighted interactive evolving neural networks are computationally equivalent to interactive Turing machines with advice, and...
متن کاملReport SYCON-91-11 ON THE COMPUTATIONAL POWER OF NEURAL NETS
This paper deals with the simulation of Turing machines by neural networks. Such networks are made up of interconnections of synchronously evolving processors, each of which updates its state according to a \sigmoidal" linear combination of the previous states of all units. The main result states that one may simulate all Turing machines by nets, in linear time. In particular, it is possible to...
متن کاملContinual Learning through Evolvable Neural Turing Machines
Continual learning, i.e. the ability to sequentially learn tasks without catastrophic forgetting of previously learned ones, is an important open challenge in machine learning. In this paper we take a step in this direction by showing that the recently proposed Evolving Neural Turing Machine (ENTM) approach is able to perform one-shot learning in a reinforcement learning task without catastroph...
متن کاملSelf-Organizing Topology Evolution of Turing Neural Networks
We present Turing's neural-network-like structures (unorganized machines) and compare them to Kau man's random boolean networks (RBN). Some characteristics of attractors are brie y presented. We then apply a self-organizing topology evolving algorithm to Turing's networks and show that the network evolves towards an average connectivity of KC = 2 for large systems (N !1).
متن کامل