TU Berlin

Service-centric NetworkingData Science

Page Content

to Navigation

Data Science

Lupe

Data in today’s business landscape are created and stored at exponentially large scales. Therefore, the need to improve business operations through data-driven decisions has emerged as an important objective for many growing companies. The field of data science addresses those needs by combining computer science, engineering, mathematics, statistics, and predictive modeling to generate analytical insights about data from a variety of sources.   Data often require a great amount of cleaning and pre-processing; so many research topics are also directed toward identifying different solutions for parallelized and distributed computing and data storage: from the big players in this market like Apache Hadoop and Spark through to CUDA. The field of data science interfaces with a variety of other disciplines, and by utilizing new computational technologies together with statistics and predictive modeling we strive to provide unique analytical insights from data at large scales.

Data Science research projects at SNET currently investigate data from the automotive, energy, and mobile communications domains. As such, we are often involved with the processing and analysis of geospatial data with both structured and unstructured formats.  Our goal is to discover the statistical relationships buried deep within data, and to use that knowledge as the framework for prototype development.

Publications

Overview of JavaScript Engines for Resource-Constrained Microcontrollers
Citation key Grun2006:Overview
Author Grunert, K.
Title of Book 5th International Conference on Smart and Sustainable Technologies 2020(SpliTech 2020)
Pages 1-7
Year 2020
DOI http://dx.doi.org/10.23919/SpliTech49282.2020.9243749
Address Split, Bol, Croatia
Abstract IoT devices often use small, constrained microcontrollers to implement their functionality. Usually, they are programmed with languages like C or C++, but there is a trend to use interpreted languages for this task. In this paper, we focus on JavaScript. We discuss the advantages and disadvantages of using this interpreted language for microcontroller development, we give an overview of the available JavaScript engines for constrained devices, and we compare these interpreters' general properties. One finding shows that the projects can be divided into embeddable interpreters (JerryScript, Duktape, mJS) and standalone runtimes (Espruino, Moddable, Mongoose OS). It was also identified that the interpreters follow different architectural approaches.
Bibtex Type of Publication SNET Data Social Ubiquitous
Link to publication Download Bibtex entry

Navigation

Quick Access

Schnellnavigation zur Seite über Nummerneingabe