How Data Science can help predict better and avoid Trump surprise?

Nov 15, 2016 posted by CircleRAM
Share

 

Data science lets you define and use information processing tools, combining speed with accuracy to reliably predict outcomes, like Trump win.

Conventional Polling Techniques cannot do the accurate prediction

Predicting human behavior is challenging and even more difficult with several factors in the mix such as location, ethnicity, income, behavior. Our national surveys are meaningless when candidates are relatively close with overlapping margins. Sampling the right people to represent entire population is complex. Not all segment of the population are equally accessible for survey. Researchers also trying to predict election outcomes are really dealing with predicting not one but two behaviors, showing up to vote and choosing a candidate. And also predicting with state by state surveys.  

Next Election Prediction

Putting Big Data information into the mix would allow highly granular segmentation of data. Slicing data by location, ethnicity, income, behavior, etc would yield a far better understanding of data. Combing this with data from search hits and social media messaging would give various ways to analyze and write models or algorithms will definitely increase the precision. And with speed and accuracy, results can be predicted instantly upto the minute.

Data Science to the rescue

Simply put, data scientists apply powerful tools and advanced statistical modeling techniques to make predictions very close to the outcomes. Note that Big Data is not a Data Science project, rather data must be operationlized in specific way to enable the prediction close to the outcome. This requires that someone continuously seek the leading edge in terms of performance and constantly rethink what's possible with Big Data. Data Science allows you to experiment with intelligence gathering technologies, developing sophisticated models and algorithms and combining disparate data sets. And it lets you ask improbable seeming questions leading the deepest data-diving expeditions and boldest explorations. Data Science is the just the beginning that will prevail decades to come.

 

 
Available Blogs
 
Nov 22, 2016 posted by CircleRAM
Nov 15, 2016 posted by CircleRAM
Aug 03, 2016 posted by CircleRAM
Mar 02, 2016 posted by CircleRAM
Feb 29, 2016 posted by CircleRAM
Feb 29, 2016 posted by CircleRAM
Feb 26, 2016 posted by CircleRAM
Feb 13, 2016 posted by CircleRAM
Feb 09, 2016 posted by CircleRAM