Artificial neural networks (ANNs) are applied to many real-world problems, ranging from pattern classification to robot control. In order to design a neural network for a particular task, the choice of an architecture (including the choice of a neuron model), and the choice of a learning algorithm have to be addressed. Evolutionary search methods can provide an automatic solution to these problems. New insights in both neuroscience and evolutionary biology have led to the development of increasingly powerful neuroevolution techniques over the last decade. This paper gives an overview of the most prominent methods for evolving ANNs with a special focus on recent advances in the synthesis of learning architectures.

An electrostatically driven silicon micro scanning mirror (MSM) for one-dimensional (1-D) and two-dimensional (2-D) deflection of light is presented. A special configuration of the driving electrodes allows the use of small electrode gaps without restricting the deflection of the plate geometrically. In this paper, the starting of the oscillation and the operation of the scanner is discussed. Experimental results show that scan angles of up to 60 degrees can be achieved at a driving voltage of only 20 V. The 2-D deflection of a laser beam is obtained by a gimbal mounting of the mirror plate. For the fabrication of the devices, SOI-wafers are used as the base material. The mechanical structures are defined by a deep silicon etch. For the electrical isolation of areas on the movable frame, polysilicon-filled trenches are used. The mechanical stability of the scanners is tested. The devices resist shocks of more than 1000 g and show no change of the resonance frequency even after long run tests of 7x10/sup 9/ periods

Abstract-Environments with varying reward contingencies constitute a challenge to many living creatures. In such conditions, animals capable of adaptation and learning derive an advantage. Recent studies suggest that neuromodulatory dynamics are a key factor in regulating learning and adaptivity when reward conditions are subject to variability. In biological neural networks, specific circuits generate modulatory signals, particularly in situations that involve learning cues such as a reward or novel stimuli. Modulatory signals are then broadcast and applied onto target synapses to activate or regulate synaptic plasticity.Artificial neural models that include modulatory dynamics could prove their potential in uncertain environments when online learning is required. However, a topology that synthesises and delivers modulatory signals to target synapses must be devised. So far, only handcrafted architectures of such kind have been attempted. Here we show that modulatory topologies can be designed autonomously by artificial evolution and achieve superior learning capabilities than traditional fixed-weight or Hebbian networks. In our experiments, we show that simulated bees autonomously evolved a modulatory network to maximise the reward in a reinforcement learning-like environment.

Autonomous car racing is a major challenge in robotics. It raises fundamental problems for classical approaches such as planning minimum-time trajectories under uncertain dynamics and controlling the car at the limits of its handling. Besides, the requirement of minimizing the lap time, which is a sparse objective, and the difficulty of collecting training data from human experts have also hindered researchers from directly applying learning-based approaches to solve the problem. In the present work, we propose a learning-based system for autonomous car racing by leveraging a high-fidelity physical car simulation, a course-progress proxy reward, and deep reinforcement learning. We deploy our system in Gran Turismo Sport, a world-leading car simulator known for its realistic physics simulation of different race cars and tracks, which is even used to recruit human race car drivers. Our trained policy achieves autonomous racing performance that goes beyond what had been achieved so far by the built-in AI, and at the same time, outperforms the fastest driver in a dataset of over 50,000 human players.

No abstract