ALL >> Business >> View Article
Data Science Institute In Hyderabad
What isData Science? The definition is still evolving and an Internet search for the term reveals dozens ofvariations. As a simple working definition, we define Data Science simply to bethe science of extractingknowledge from data.
From the recent attention information Science has received in tutorial journals and thepopular press, one gets the distinct impression that this is often a brand new discipline.
But is it really? Experts in dataanalysis, most notably statisticians, have been extracting knowledge from data for decades. In a recent articleinForbesentitled “A Very Short History of Data Science,”1Gil Press traces the origins of Data Science as adiscipline back to an article by John Tukey in 1962 called “The Future of Data Analysis”2in which he wrote“Data analysis, and the parts of statistics which adhere to it, must. . .
take on the charac-teristics of science instead of those of arithmetic.
. . data analysis is intrinsically anempirical science. . . How vital and how important. . . is the rise of the stored-programelectronic computer?
In several instances the solution could surprise several by being ‘impor-tant however not very important,’ although in others there is no doubt but what the computer has been‘vital.’”Given this early
recognition by Tukey and others of the importance of Data Science as a field distinctfrom statistics, why has it taken so long for it to be recognized as a pronounced and important discipline?The most likely answer is that it was several more decades before the confluence of computational methods,computing technology, and mathematical techniques that allowed Tukey’s vision to be realized would occur.Although it was possible to envision modern Data Science several decades ago, we simply did not have themeans to generate, store, and share the volumes of data required for many of the applications that are drivingmodern needs and trends.Big DataAnother term that has recently gained traction and cachet in both the popular press and academic circles isBig Data. It is clear that we have now entered “the age of Big Data” and much of the recent emphasis onData Science has been borne out of the explosion in the availability of Big Data, usually described as datahaving the following characteristics3:•Volume.It is estimated that tens of exabytes of data are gathered worldwide each day and this amountis forecasted to double every 40 months. For example, it is estimated that Walmart collects more than 2petabytes of data every hour from its customer transactions.•Velocity.For many applications, the speed of data creation is even more important than its volume.Real-time information can help companies be more agile than their competitors.•Variety.Big Data includes a wide variety of data types, including Facebook statuses, pictures onGoogle’s Picasa or Flickr, articles in Wikipedia, Tweets on Twitter, readings from various sensors,YouTube movies, and much more. All of these are sources of unstructured data, not suitable to bestored in classical relational databases, which assume that data possess a certain structure.It would be a mistake, however, to equate Data Science with Big Data. Data does not have to be “big” inorder for the extraction of knowledge from it to be challenging.AnalyticsAnalyticsis another term that has been variously defined and has recently increased in usage and popularity.The Institute for Operations Research and Management Science (INFORMS), the leading professional societyof Analytics experts, defines it as thescientific process of transforming data into insight for makingbetter decisions.4This definition differs from that of Data Science in that it makes explicit the end goalof havingthe insight to make an informed decision. The data is one input into a cyclic process, shown inFigure 1(a), in which the collection of data drives decisions, which in turn drive the collection of more data.Although it is easy to collect a large volume of data without first thinking about what decisions these datawill be used to make, this indiscriminate approach collection is not likely to lead to meaningful results. Thecyclic nature of the Analytics process is critical.In “A Taxonomy of Data Science,”5Mason and Wiggins provide an alternative view of this processand state that there are five steps data
scientists follow in analyzing data: get, Scrub, Explore, Model,and Interpret.
This describes a similar cycle, but explicitly includes the concept of developing a “model”following the exploration phase. Exploration can be seen as an informal and usually human-driven and with deliveries made from a single warehouse, this problem (knownin academic papers as theVehicle Routing Problem) is difficult to solve.Big Data AnalyticsWhen problems that are already computationally difficult at a small scale are made more realistic by includingfine-grained data (e.g., demand forecasts) and the problem is scaled to the size faced by a company suchas Amazon or Google, then we have entered the realm ofBig Data Analytics. The techniques of Big DataAnalytics encompass the computational challenges involved both in theanalysisof the data and in theexploitationof it as part of a data-driven decision-making process. Simply put, Big Data Analytics (seeFigure 1(b)) is the confluence of Big Data, Big Analytics, and Big Computation.Machine learning, data mining, social network analysis, financial optimization, healthcare analytics, andcomputational biology are some of many prominent application domains where Big Data is available andmathematical optimization modeling is the natural framework for making decisions. In these applications,decision problems with millions or billions of variables are commonplace. Classical optimization algorithmsare not designed to scale to instances of this size. There is a need to continually develop new approaches.ISE and Big Data AnalyticsBeginning around the year 2000, ISE instituted a departmental focus on Analytics. Since that time, Analyticshas been our core strength and has since been identified in our strategic plan as both our primary strategicfocus and our primary growth area. Throughout, we have grown our expertise in this area through neweducational programs and initiatives, targeted hiring, and the development of new labs and research centers.Among the educational initiatives that facilitated this growth was the development of a new undergraduateprogram calledInformation and Systems Engineering(I&SE).
Business Articles1. What Is Hemp Protein And How Can It Help You?
Author: Unknown Member
2. Important For Epoxy Coatings
Author: Unknown Member
3. Java Development Training And Jobs
Author: Zain Malik
4. All You Need To Know About Iso 13485 Auditor Training And Its Requirements
Author: Charles Wilson
5. Is Bitdefender Antivirus Plus Right For You?
6. Beneficial Tips To Consider While Going On Hunting
Author: Unknown Member
7. Lead Generation Agency In Mumbai
8. Know What A Right Translation Agency Can Do For You
9. Keuntungan Dari Perjalanan Bus
Author: Unknown Member
10. How To Recycle Moving Cardboard Boxes?
Author: Jeff Parker
11. Get The Best Employee Hiring, Human Resource And Business Plan Template For Your Business
Author: Diptiranjan Pradhan
12. It’s Mandatory To Select A Reputable Ferroalloy Supplier
13. Applications Of Steel In Different Fields
14. Different Types Of Ferroalloys And Their Uses
15. Look For A Trusted And Experienced Ferroalloy Supplier