Articles

Distributed learning in large-scale multi-agent games: A modified fictitious play approach

Swenson B., Kar S., Xavier J.

Conference Record - Asilomar Conference on Signals, Systems and Computers

2012

Abstract:

The paper concerns the development of distributed equilibria learning strategies in large-scale multi-agent games with repeated plays. With inter-agent information exchange being restricted to a preassigned communication graph, the paper presents a modified version of the fictitious play algorithm that relies only on local neighborhood information exchange for agent policy update. Under the assumption of identical agent utility functions that are permutation invariant, the proposed distributed algorithm leads to convergence of the networked-averaged empirical play histories to a subset of the Nash equilibria, designated as the consensus equilibria. Applications of the proposed distributed framework to strategy design problems encountered in large-scale traffic networks are discussed.

Distributed Linear Estimation Via a Roaming Token

Balthazar L., Xavier J., Sinopoli B.

IEEE Transactions on Signal Processing

2020

Abstract:

We present an algorithm for the problem of linear distributed estimation of a parameter in a network where a set of agents are successively taking measurements. The approach considers a roaming token in a network that carries the estimate, and jumps from one agent to another in its vicinity according to the probabilities of a Markov chain. When the token is at an agent it records the agent’s local information. We analyze the proposed algorithm and show that it is consistent and asymptotically optimal, in the sense that its mean-square-error (MSE) rate of decay approaches the centralized one as the number of iterations increases. We show these results for a scenario where the network changes over time, and we consider two different sets of assumptions on the network instantiations: (I) they are i.i.d. and connected on the average, or (II) that they are deterministic and strongly connected for every finite time window of a fixed size. Simulations show our algorithm is competitive with consensus+innovations and diffusion type of algorithms, achieving a smaller MSE at each iteration in all considered scenarios.

Distributed Nesterov gradient methods for random networks: Convergence in probability and convergence rates

Jakovetic D., Xavier J., Moura J.M.F.

ICASSP, IEEE International Conference on Acoustics, Speech and Signal Processing - Proceedings

2014

Abstract:

We consider distributed optimization where N nodes in a generic, connected network minimize the sum of their individual, locally known, convex costs. Existing literature proposes distributed gradient-like methods that are attractive due to computationally cheap iterations and provable resilience to random inter-node communication failures, but such methods have slow theoretical and empirical convergence rates. Building from the centralized Nesterov gradient methods, we propose accelerated distributed gradient-like methods and establish that they achieve strictly faster rates than existing distributed methods. At the same time, our methods maintain cheap iterations and resilience to random communication failures. Specifically, for convex, differentiable local costs with Lipschitz continuous and bounded derivative, we establish (with respect to the cost function optimality) convergence in probability and convergence rates in expectation and in second moment.

Distributed Nesterov-like gradient algorithms

Jakovetic D., Moura J.M.F., Xavier J.

Proceedings of the IEEE Conference on Decision and Control

2012

Abstract:

In classical, centralized optimization, the Nesterov gradient algorithm reduces the number of iterations to produce an ε-accurate solution (in terms of the cost function) with respect to ordinary gradient from O(1/ε) to equation. This improvement is achieved on a class of convex functions with Lipschitz continuous first derivative, and it comes at a very small additional computational cost per iteration. In this paper, we consider distributed optimization, where nodes in the network cooperatively minimize the sum of their private costs subject to a global constraint. To solve this problem, recent literature proposes distributed (sub)gradient algorithms, that are attractive due to computationally inexpensive iterations, but that converge slowly-the ε error is achieved in O(1/ε 2 ) iterations. Here, building from the Nesterov gradient algorithm, we present a distributed, constant step size, Nesterov-like gradient algorithm that converges much faster than existing distributed (sub)gradient methods, with zero additional communications and very small additional computations per iteration k. We show that our algorithm converges to a solution neighborhood, such that, for a convex compact constraint set and optimized stepsize, the convergence time is O(1/ε). We achieve this on a class of convex, coercive, continuously differentiable private costs with Lipschitz first derivative. We derive our algorithm through a useful penalty, network’s Laplacian matrix-based reformulation of the original problem (referred to as the clone problem) – the proposed method is precisely the Nesterov-gradient applied on the clone problem. Finally, we illustrate the performance of our algorithm on distributed learning of a classifier via logistic loss.

Distributed Optimization with Local Domains: Applications in MPC and Network Flows

Mota J.F.C., Xavier J.M.F., Aguiar P.M.Q., Puschel M.

IEEE Transactions on Automatic Control

2015

Abstract:

We consider a network where each node has exclusive access to a local cost function. Our contribution is a communication-efficient distributed algorithm that finds a vector x* minimizing the sum of all the functions. We make the additional assumption that the functions have intersecting local domains, i.e., each function depends only on some components of the variable. Consequently, each node is interested in knowing only some components of x*, not the entire vector. This allows improving communication-efficiency. We apply our algorithm to distributed model predictive control (D-MPC) and to network flow problems and show, through experiments on large networks, that the proposed algorithm requires less communications to converge than prior state-of-the-art algorithms.

Distribution grid reconfiguration reduces power losses and helps integrate renewables

Lueken C., Carvalho P.M.S., Apt J.

Energy Policy

2012

Abstract:

A reconfigurable network can change its topology by opening and closing switches on power lines. We use real wind, solar, load, and cost data and a model of a reconfigurable distribution grid to show that reconfiguration allows a grid operator to reduce operational losses as well as to accept more intermittent renewable generation than a static configuration can. Net present value analysis of automated switch technology shows that the return on investment is negative for this test network when considering only loss reduction, but that the investment is attractive under certain conditions when reconfiguration is used to minimize curtailment.

Diversification by young, small firms: the role of pre-entry resources and entry mistakes

Baptista R, MKaraö M., Leitão J.C.

Small Business Economics

2019

Abstract:

We investigate the determinants of young, small firm diversification by using longitudinal linked employer-employee data. We focus particularly on the role played by the sharing of managerial and qualified human resources, as well as market uncertainty and entry mistakes. We find that a small but significant proportion of young, small firms diversify in their first years. Firms with a greater proportion of managers and qualified human resources are more likely to diversify early, lending credence to the resource-based view of diversification. Firms entering volatile markets are more likely to diversify earlier as well, suggesting that entry mistakes and escape from uncertain, Schumpeterian environments also influence diversification. The inspection of survival patterns of diversified firms sheds further light on the importance of these two determinants of diversification. Keywords Diversification Start-ups Small business Firm resources Uncertainty Entry mistakes

Do entry regulations deter entrepreneurship and job creation? Evidence from recent reforms in Portugal

Branstetter L., Lima F., Taylor L.J., Venancio A.

Economic Journal

2014

Abstract:

We evaluate the consequences of a recent regulatory reform in Portugal, which substantially reduced the cost of firm entry. Our analysis uses matched employer–employee data, which provide unusually rich information on the characteristics of founders and employees associated with new firms before and after the reform. We find that the short‐term consequences of the reform were as one would predict with a standard economic model of entrepreneurship: the reform resulted in increased firm formation and employment, but mostly among ‘marginal firms’ that would have been most readily deterred by existing heavy entry regulations. These marginal firms were typically small, owned by relatively poorly educated entrepreneurs, and operating in low‐technology sectors (agriculture, construction and retail trade). In comparison to firms that entered in the absence of the reform, these marginal firms were less likely to survive their first two years.

Domiciliary Hospitalization through Wearable Biomonitoring Patches: Recent Advances, Technical Challenges, and the Relation to Covid-19

Silva A., Tavakoli M.

Sensors

2020

Abstract:

This article reviews recent advances and existing challenges for the application of wearable bioelectronics for patient monitoring and domiciliary hospitalization. More specifically, we focus on technical challenges and solutions for the implementation of wearable and conformal bioelectronics for long-term patient biomonitoring and discuss their application on the Internet of medical things (IoMT). We first discuss the general architecture of IoMT systems for domiciliary hospitalization and the three layers of the system, including the sensing, communication, and application layers. In regard to the sensing layer, we focus on current trends, recent advances, and challenges in the implementation of stretchable patches. This includes fabrication strategies and solutions for energy storage and energy harvesting, such as printed batteries and supercapacitors. As a case study, we discuss the application of IoMT for domiciliary hospitalization of COVID 19 patients. This can be used as a strategy to reduce the pressure on the healthcare system, as it allows continuous patient monitoring and reduced physical presence in the hospital, and at the same time enables the collection of large data for posterior analysis. Finally, based on the previous works in the field, we recommend a conceptual IoMT design for wearable monitoring of COVID 19 patients.