This site uses cookies.
Some of these cookies are essential to the operation of the site,
while others help to improve your experience by providing insights into how the site is being used.
For more information, please see the ProZ.com privacy policy.
This person has a SecurePRO™ card. Because this person is not a ProZ.com Plus subscriber, to view his or her SecurePRO™ card you must be a ProZ.com Business member or Plus subscriber.
Affiliations
This person is not affiliated with any business or Blue Board record at ProZ.com.
Services
Translation
Expertise
Specializes in:
Computers: Systems, Networks
Electronics / Elect Eng
Computers: Software
Also works in:
Computers: Hardware
Engineering (general)
Computers (general)
Automation & Robotics
More
Less
Rates
English to Spanish - Standard rate: 0.15 EUR per word / 20 EUR per hour Spanish to English - Standard rate: 0.12 EUR per word / 12 EUR per hour
Payment methods accepted
PayPal, Money order
Portfolio
Sample translations submitted: 1
English to Spanish: Real-time information processing of environmental sensor network data using Bayesian Gaussian processes General field: Tech/Engineering
Source text - English 1. INTRODUCTION
Sensor networks have recently generated a great deal of research interest within
the computer and physical sciences, and their use for the scientific monitoring of
remote and hostile environments is increasingly common-place. While early sensor
networks were a simple evolution of existing automated data loggers that collected
data for later off-line scientific analysis, more recent sensor networks now make
their data available in real-time through the internet, and increasingly perform
some form of real-time data processing or aggregation to provide useful summary
information to users [Hart and Martinez 2006].
Providing real-time access to sensor data in this way presents many novel challenges; not least the need for self-describing data formats, and standard protocols
such that the existence and capabilities of sensors can be advertised to potential users. However, more significantly for us, many of the information processing tasks that would previously have been performed off-line by the expert owner
or single user of an environmental sensor network (such as detecting faulty sensors, performing ‘data cleaning’ to remove erroneous readings, fusing noisy measurements from several sensors, and deciding how frequently data should be collected), must now be performed autonomously in real-time. Furthermore, since
such information processing is likely to be performed within centralised sensor
repositories — of which Weather Underground (http://www.wunderground.com,
pachube (http://www.pachube.com/) and Microsoft Research’s SensorMap (http:
//atom.research.microsoft.com/sensormap/) are embryonic examples — and
will be applied to open sensor networks where additional sensors may be deployed
at any time, and existing sensors may be removed, repositioned or updated after deployment (such as the rooftop weather sensors within the Weather Underground),
these information processing tasks may have to be performed with only limited
knowledge of the precise location, reliability, and accuracy of each sensor.
Now, many of the information processing tasks described above have previously
been tackled by applying principled Bayesian methodologies from the academic
literature of geospatial statistics and machine learning: specifically, kriging [Cressie
1991] and Gaussian processes [Rasmussen and Williams 2006]. However, due to
the computational complexity of these approaches, to date they have largely been
used off-line in order to analyse and re-design existing sensor networks (e.g. to
reduce maintenance costs by removing the least informative sensors from an existing
sensor network [Fuentes et al. 2007], or to find the optimum placement of a small
number of sensors, after a trial deployment of a larger number has collected data
indicating their spatial correlation [Krause et al. 2006]). Alternatively, they have
been applied to single sensors and have ignored the correlations that exist between
different sensors within the network [Kho et al. 2009; Padhy et al. 2010]. Thus,
there is a clear need for more computationally efficient algorithms, in order that
this information processing can be performed at scale in real-time.
Against this background, this paper describes our work developing just such an
algorithm. More specifically, we present a novel iterative formulation of a multioutput Gaussian process (GP) that uses a computationally efficient implementation
of Bayesian Monte Carlo to marginalise hyperparameters, efficiently re-uses previous computations by following an online update procedure as new data sequentially
ACM Transactions of Sensor Networks, Vol. V, No. N, Month 20YY.
Translation - Spanish INTRODUCCION
Las redes de sensores han generado recientemente un gran interés por
parte de investigadores informáticos y físicos y su uso para la
monitorización científica de entornos hostiles y remotos es cada vez más
común. Al igual que las primeras redes de sensores eran una simple
evolución de los registradores de datos automáticos ya existente que
recogen datos para su posterior análisis científico sin conexión, las redes
de sensores más recientes generalmente hacen posible que los datos
que se están usando estén disponibles en internet. Por tanto, se usan
cada vez más para el seguimiento a tiempo real de acontecimientos
medioambientales tales como inundaciones o tormentas (véase nota 7
ara más información acerca de este tipo de redes de sensores
medioambientales)
Tal acceso en tiempo real a los datos del sensor es también una
característica de los sistemas de sensores extendidos en el que los
sensores son propiedad de múltiples tipos de accionistas (por ejemplo
particulares, propietarios de edificios y autoridades públicas) son
desplegados en todas partes dentro de los entornos urbanos y pone
disponible su información a múltiples usuarios a través de interfaces
estándar. (Véase el proyecto CitySense de la universidad de Harvard y
el proyecto SenseWeb de Microsoft). Tales redes tienen muchas
aplicaciones, incluidas la monitorización del tráfico o la contaminación,
y dentro del proyecto ALADDIN (http://www.aladdinproject.org),
estamos buscando utilizar tales redes para proveer soporte de
More
Less
Experience
Years of experience: 1. Registered at ProZ.com: Sep 2013.