1,722,416 research outputs found
Sparse Distributed Learning Based on Diffusion Adaptation
No full textThis article proposes diffusion LMS strategies for distributed estimation over adaptive networks that are able to exploit sparsity in the underlying system model. The approach relies on convex regularization, common in compressive sensing, to enhance the detection of sparsity via a diffusive process over the network. The resulting algorithms endow networks with learning abilities and allow them to learn the sparse structure from the incoming data in real-time, and also to track variations in the sparsity of the model. We provide convergence and mean-square performance analysis of the proposed method and show under what conditions it outperforms the unregularized diffusion version. We also show how to adaptively select the regularization parameter. Simulation results illustrate the advantage of the proposed filters for sparse data recovery.AS
Sparse distributed learning based on diffusion adaptation
No full textThis article proposes diffusion LMS strategies for distributed estimation over adaptive networks that are able to exploit sparsity in the underlying system model. The approach relies on convex regularization, common in compressive sensing, to enhance the detection of sparsity via a diffusive process over the network. The resulting algorithms endow networks with learning abilities and allow them to learn the sparse structure from the incoming data in real-time, and also to track variations in the sparsity of the model. We provide convergence and mean-square performance analysis of the proposed method and show under what conditions it outperforms the unregularized diffusion version. We also show how to adaptively select the regularization parameter. Simulation results illustrate the advantage of the proposed filters for sparse data recovery
Distributed Detection over Adaptive Networks: Refined Asymptotics and the Role of Connectivity
No full textWe consider distributed detection problems over adaptive networks, where dispersed agents learn continually from streaming data by means of local interactions. The requirement of adaptation allows the network of detectors to track drifts in the underlying hypothesis. The requirement of cooperation allows each agent to deliver a performance superior to what would be obtained if it were acting individually. The simultaneous requirements of adaptation and cooperation are achieved by employing diffusion algorithms with constant step-size μ. By conducting a refined asymptotic analysis based on the mathematical framework of exact asymptotics, we arrive at a revealing understanding of the universal behavior of distributed detection over adaptive networks: as functions of 1/μ, the error (log-)probability curves corresponding to different agents stay nearly-parallel to each other, however, these curves are ordered following a criterion reflecting the degree of connectivity of each agent. Depending on the combination weights, the more connected an agent is, the lower its error probability curve will be. The analysis provides explicit analytical formulas for the detection error probabilities and these expressions are also verified by means of extensive simulations. We further enlarge the reference setting from the case of doubly-stochastic combination matrices considered in [1] and [2] to the more general and demanding setting of right-stochastic combination matrices; this extension poses new and interesting questions in terms of the interplay between the network topology, the combination weights, and the inference performance. The potential of the proposed methods is illustrated by application of the results to canonical detection problems, to typical network topologies, for both doubly-stochastic and right-stochastic combination matrices. Interesting and somehow unexpected behaviors emerge, and the lesson learned is that connectivity matters.AS
Detection over Diffusion Networks: Asymptotic Tools for Performance Prediction and Simulation
No full textExploiting recent progress [1]-[4] in the characterization of the detection performance of diffusion strategies over adaptive multi-agent networks: i) we present two theoretical approximations, one based on asymptotic normality and the other based on the theory of exact asymptotics; and ii) we develop an efficient simulation method by tailoring the importance sampling technique to diffusion adaptation. We show that these theoretical and experimental tools complement each other well, with their combination offering a substantial advance for a reliable quantitative detection-performance assessment. The analysis provides insight into the interplay between the network topology, the combination weights, and the inference performance, revealing the universal behavior of diffusion-based detectors over adaptive networks.AS
Google trends and cryptocurrencies: A nonparametric causality-in-quantiles analysis
No full textPurpose: This article aims to uncover the impact of Google Trends on cryptocurrency markets beyond Bitcoin during the time of increased attention to altcoins, especially during the COVID-19 pandemic. Design/methodology/approach: This paper analyses the nexus among the Google Trends and six cryptocurrencies, namely Bitcoin, New Economy Movement (NEM), Dash, Ethereum, Ripple and Litecoin by utilizing the causality-in-quantiles technique on data comprised of the years January 2016–March 2021. Findings: The findings show that Google Trends cause the Litecoin, Bitcoin, Ripple, Ethereum and NEM prices at majority of the quantiles except for Dash. Originality/value: The findings will help investors to develop more in-depth understanding of impact of Google Trends on cryptocurrency prices and build successful trading strategies in a more matured digital assets ecosystem.</p
Diffusion-based adaptive distributed detection: Steady-state performance in the slow adaptation regime
Get PDFThis paper examines the close interplay between cooperation and adaptation for distributed detection schemes over fully decentralized networks. The combined attributes of cooperation and adaptation are necessary to enable networks of detectors to continually learn from streaming data and to continually track drifts in the state of nature when deciding in favor of one hypothesis or another. The results in this paper establish a fundamental scaling law for the steady-state probabilities of miss detection and false alarm in the slow adaptation regime, when the agents interact with each other according to distributed strategies that employ small constant step-sizes. The latter are critical to enable continuous adaptation and learning. This paper establishes three key results. First, it is shown that the output of the collaborative process at each agent has a steady-state distribution. Second, it is shown that this distribution is asymptotically Gaussian in the slow adaptation regime of small step-sizes. Third, by carrying out a detailed large deviations analysis, closed-form expressions are derived for the decaying rates of the false-alarm and miss-detection probabilities. Interesting insights are gained from these expressions. In particular, it is verified that as the step-size decreases, the error probabilities are driven to zero exponentially fast as functions of , and that the exponents governing the decay increase linearly in the number of agents. It is also verified that the scaling laws governing the errors of detection and the errors of estimation over the network behave very differently, with the former having exponential decay proportional to , while the latter scales linearly with decay proportional to . Moreover, and interestingly, it is shown that the cooperative strategy allows each agent to reach the same detection performance, in terms of detection error exponents, of a centralized stochastic-gradient solution. The results of this paper are illustrated by applying them to canonical distributed detection problems
Bloom Filter-Based Content Discovery and Retrieval for Information-Centric Networks
Get PDFNamed Data Networking (NDN) requires routing protocols that use content object names for routing clients’ requests. In this thesis, we develop routing protocols for NDN based on content advertisements that we compress using Bloom filters (BF). We propose push-based Bloom Filter-based Routing (BFR) and pull-based BFR for NDN. Push-based BFR advertises all provided content object names, while pull-based BFR only advertises the requested content object names. Therefore, pull-based BFR outperforms push-based BFR in terms of the required communication and storage overhead for content advertisements. To reduce content retrieval delay, we propose to use Network Coding (NC)-based content retrieval. We use the BF-based information distributed for content discovery to select network codes. The proposed NC-based protocol uses a constraint on the equation system size and BF-based feedbacks to control codeblock size. We show that the proposed NC-based protocol achieves lower average content block retrieval delay than push-based and pull-based BFR. Service-Centric Networking [21] requires load balancing mechanisms to route service requests. To address this requirement, Layered-Service Centric Networking (L-SCN) [31] proposed to divide nodes into domains where each domain of nodes is managed by a supernode. However, L-SCN lacks algorithms to select supernodes in the network topology. We present supernode selection algorithms based on the construction of Dominating Sets (DS) and Connected Dominating Sets (CDS) over the network topology. Then, we propose intra-domain and inter-domain BF-based routing protocols for routing service requests. We show that our CDS-based routing protocols require much less bandwidth overhead for routing than both DS-based routing and NDN multicast strategy. Further, we show that both DS-based and CDS-based routing protocols achieve significantly less service retrieval delay than the NDN multicast strategy
Mapping the Void: Landscapes of Roman Travertine Extraction
Get PDFRoman Travertine is the stone of the empire, it has
characterised the city of Rome, and unexpectedly
conditioned the Tiburtine hinterlands. This thesis
reveals the hidden scars and enduring legacies of
material extraction on landscapes, centering on the
ancient yet persistently exploited Roman travertine
quarries near Rome. Unfolding the tension between
rugged extraction sites and polished urban spaces, the
study critiques architecture’s reliance on a “Western
Paradox” where development and destruction
coexist. It places the landscape discinpline right in
the middle, an accomplice and an active actor in the
consequences of material consumption, attempting to
answer the question: How can the landscape discipline
represent the ecological and social conflicts caused
by resource extraction, and what impact does
extraction have on how landscapes are understood?
Through political ecology and interdisciplinary
analysis, it unearths the rich ecological, historical,
and political narratives surrounding travertine,
attempting to reveal its consequences through a
mapping project. In three sections, the research
first establishes the theoritical framework, addresses
Necropolitics of Extraction, exposing architecture’s
detachment from the material realities of production
and spotlighting the social and environmental
costs of material consumption. The second section,
Lapis Tiburtinus, turns to travertine itself, tracing
its symbolic and structural roles across millennia,
from shaping Rome’s grandeur to a globalized
commodity that transformed its landscapes. The final
section, Mapping Travertine Extraction, uses dynamic
cartographies of water, waste, and labour to bring the
invisible forces of extraction into stark relief, illustrating
both local devastations and global dependencies
Design in the Meta.space: A speculative investigation into the spatial perception at the virtual-analog frontier
No full textThe concept of the metaverse, a virtual world that runs parallel to our own and represents the next stage of the internet, has sparked intense debates since the blockchain craze began. However, the ability to construct and shape worlds by manipulating social, real, and imaginary spaces has arguably been a crucial aspect of artistic and scientific creativity throughout history. Accompanying an exhibition at the Francisco Carolinum in Linz, the scientific anthology Meta.space—Visions of Space from the Middle Ages to the Digital Age explores this multifaceted history from an art-historical perspective. Featuring seventy-five positions in analog, digital, and inter-media art, it provides a foundational contribution to ongoing discussions about the metaverse. The anthology covers a range of topics, including early painterly solutions to pictorial space problems in the fifteenth century, how works of sculpture create spaces, and the sensory, scholarly, and technological practices involved in charting spaces. Additionally, the book critically examines the dystopian and utopian potential of cutting-edge metaspace concepts
- …
