Thursday, October 6, 2016

Business Intelligence VS Data Science

just ask any practitioner within the field and you will notice at first, the answers appeared to be a random mix, but once they were collated and reviewed, there was a general consensus.

Some of the more in-depth answers included

“Data Science is the core for BI; it is the one in charge of developing the algorithms and strategies to data patterns and trends in data. Business Intelligence uses what Data Science produces in order to give enterprise value to its information.”

“BI is more of a decision support system which helps business drive focus while Data Science is more statistics-driven, and we need folks to pay more attention to granularities within the data. BI is more about ETL and dashboards presentation while Data Science is more focused on finding anomalies within data and establishing the pattern.”

Data Science generates insights and finds anomalies in data, scans, and predicts the future. BI helps to see where your business is at a point in time, generates reports, and dashboards. It is limited to past and present.”

“To me Data Science sounds like the study of making use of data to enable and optimize the realization of information knowledge from it. Business Intelligence makes me think of moving transactional data to a Data Warehouse and writing reports from there. But maybe that isn’t all that different.”

Type of question BI systems can answer .


  • How many customers have visited our website every day for the past five years?
  • What is product X’s revenue trend over the holidays?
  • How often do we need to replace a particular part on our assembly line?
  • What has been our most successful sales campaign based on certain metrics?
  • What is the historical water usage during summer months of our customers in a given locale?
  • Who are our most valuable customers?
  • How many page views did we get from a given article last month?
  • What is our total inventory stock of agiven item?

Type of questions Data Science allow you to answer 


  • How many customers will visit our website next month and buy something rather than just click through?
  • Based on their past behaviors, which customers are most likely to be interested in our new product Y?
  • What correlations exist among season, geographical region, certain product lines, and a given social media campaign when looking at why a customer will purchase a product versus just browse our product listings?
  • What variables will make our next marketing campaign the most successful?
  • What variables make one patient more likely to have a given medical condition over other patients?Which medical treatments and preventative measures are most successful in helping such conditions?
  • How can we use geo-location data with real-time weather patterns to figure out how best to route incoming and outgoing planes?
  • What can the combination of historical flooding data and remote maintenance sensors data tell
  • us about creating preemptive flood overflow plans for our most problematic city neighborhoods?
Over the next 5 years do you plan to invest in Predictive Analytics (BigData/Data Science)?”





Monday, June 27, 2016

Setup and Configuer R Services in SQL 2016

SQL Server R Services is designed to help you complete data science tasks.  scale analysis to billions of records without additional hardware, boost performance, and avoid unnecessary data movements.

Now you can put your R code into production without having to re-write it in another language. It also makes it easy to use R for statistical computations that might be difficult to implement using SQL. At the same time, you can leverage the power of SQL Server to achieve maximum performance, using features such as the in-memory database engine and column store indexes.

Here you will find detailed steps how you can install and configure R services and run your first code .

  1. launch the Setup of SQL Server
  2. Go to Installation clickable link in the left pane
  3. Click at the first link the right pane as below screen shoot shows


 4. on the feature selection page select these options
  •  Database Engine Services
  •  R Services (In-Database
5. On the page, Consent to Install Microsoft R Open, click Accept.
6. if you are using a server doesn't connect to internet you will be asked to provide the location of R   packages you can download from here ( https://go.microsoft.com/fwlink/?LinkId=761266&lcid=1033  https://go.microsoft.com/fwlink/?LinkId=735051&lcid=1033 )
7. point to the location of the files downloaded in step 6 and then press next
8. installation progress started
9. you should now successfully installed R service


10. connect to the instance where the R service installed
11. Run the following T-SQL Code to enable
Exec sp_configure  'external scripts enabled', 1 
Reconfigure  with  override

12. Restart the SQL Server Service (DB Engine)
13. the instance is ready to run the first R code
exec sp_execute_external_script  @language =N'R',   
@script=N'OutputDataSet<-inputdataset br="" nbsp="">@input_data_1 =N'select 1 as hello'   
with result sets (([hello] int not null));   
go
  

Enjoy :) this is the output of you first R' code