Cointegration analysis of money demand in the Euro Area

The analysis of the evolution of the M3 money aggregate is an important element in the definition and implementation of monetary policy for the ECB. A well-defined and stable long run demand function is an essential requisite for M3 to be a valid monetary tool. Therefore, this paper analyzes based in cointegration techniques the existence of a long run money demand, estimating it and testing its stability for the Euro Area and for ten of its member countries. Specifically, bearing in mind the high degree of monetary instability that the current economic crisis has created in the Euro Area, we also test whether this has had a noticeable impact in the cointegration among real money demand and its determinants. The analysis gives evidence of the existence of a long run relationship when the aggregated Euro Area and six of the ten countries are considered. However, these relationships are highly instable since the outbreak of the financial crisis, leading in some cases to even rejecting cointegration. All this suggests that the ECB’s strategy of focusing in the M3 monetary aggregates could not be a convenient approach under the current circumstances

Learning Multirobot Hose Transportation and Deployment by Distributed Round-Robin Q-Learning

Multi-Agent Reinforcement Learning (MARL) algorithms face two main difficulties: the curse of dimensionality, and environment non-stationarity due to the independent learning processes carried out by the agents concurrently. In this paper we formalize and prove the convergence of a Distributed Round Robin Q-learning (D-RR-QL) algorithm for cooperative systems. The computational complexity of this algorithm increases linearly with the number of agents. Moreover, it eliminates environment non sta tionarity by carrying a round-robin scheduling of the action selection and execution. That this learning scheme allows the implementation of Modular State-Action Vetoes (MSAV) in cooperative multi-agent systems, which speeds up learning convergence in over-constrained systems by vetoing state-action pairs which lead to undesired termination states (UTS) in the relevant state-action subspace. Each agent's local state-action value function learning is an independent process, including the MSAV policies. Coordination of locally optimal policies to obtain the global optimal joint policy is achieved by a greedy selection procedure using message passing. We show that D-RR-QL improves over state-of-the-art approaches, such as Distributed Q-Learning, Team Q-Learning and Coordinated Reinforcement Learning in a paradigmatic Linked Multi-Component Robotic System (L-MCRS) control problem: the hose transportation task. L-MCRS are over-constrained systems with many UTS induced by the interaction of the passive linking element and the active mobile robots.

On-demand generation and removal of alginate biocompatible microvalves for flow control in microfluidics

[EN] This paper describes, for the first time, the use of alginate hydrogels as miniaturised microvalves within microfluidic devices. These biocompatible and biodegradable microvalves are generated in situ and on demand, allowing for microfluidic flow control. The microfluidic devices were fabricated using an origami inspired technique of folding several layers of cyclic olefin polymer followed by thermocompression bonding. The hydrogels can be dehydrated at mild temperatures, 37◦C, to slightly open the microvalve and chemically erased using an ethylenediaminetetraacetic acid disodium salt (EDTA) solution, to completely open the channel, ensuring the reusability of the whole device and removal of damaged or defective valves for subsequent regeneration.