Hello Data Inquisitors,
Let me continue from my last blog Do-we-really-need-data-scientist.html where I wrote about how Data Scientists are different from Database Administrators. To conclude my last blog, I would say they complement and supplement each other in a big way. One will not be able to accomplish objectives without other’s existence, it is same as Hen and an Egg situation.
“Data could be of any form or quality but right inference out of it is the science. – AG”
Let us now focus and talk about what Data Scientist can do to make entities grow. They enable them with the guided and informed decisions by sharing the precise information at much need time. Basic principle how anyone can achieve goal is, by taking THE right decision at THE right time with THE right set of information. This is possible because Data Scientists enable entities with the correct visualization of the data to make accurate decision at appropriate time. Keeping in view the targets/objectives leaders can define strategy and roadmap for the growth trajectory.
How Data Scientists can contribute to the journey of success? Let me help everyone understand this by dividing my response in 2 parts covering E2E journey i.e., ENABLE and EXECUTION.
Enabling leaders with much needed statistical data by connecting the dots around raw figures. Data viewed as simple numbers are presented as dashboards by applying complex algorithms. It is a unique skill that Data Scientist possess to convert raw numbers to meaningful information. Data Scientists hold the onus of not just output but making sure input data which they consume displays right behavior. They had to labor the data through well-defined statistical route as explained below:
To start with data gets churned using Exploratory Data Analysis, which helps make sure right data is available for ingestion. From the data lake, Data Scientists must pick up the right sized sample making sure it is normalized and fit well in bell curve (preferably). Sample should reflect the right behavior based on Industry Domain, with appropriate Skewness level and defined peak to focus on.
Perform What-If analysis to identify the right parameters to concentrate on adopting Simulation Techniques. Sensitivity analysis helps identify critical paths.
Data and Text mining is carried out to map top 20% of scenarios leading to 80% of scenarios, by forming a Data Word Cloud.Validating the expected outcome during and post Execution holds the ground right that actions identified and executed are appropriate for the results.
Hypothesis Analysis/testing is carried out to understand the confidence level of meeting the objectives post correction.
Derive the right trends against Mean, Median, Mode and Standard Deviations to capture the normalized behavior of the outcome data.
Control charts are draw to monitor variations against defined limits of success quantitatively.
“Talented data scientists leverage data that everybody sees; visionary data scientists leverage data that nobody sees.” ―Vincent Granville, Executive Data Scientist & Co-Founder at Data Science Central
Looking at the data differently and make it a valuable statistical figure is the key skill for Data Scientist. Data Scientists role in E2E journey start from day Zero where it helps define Data and it properties.
In my next blog, I will focus more on how Data Scientist leverage their analytical skills and tools to be more accurate and precise in their outcome. Transforming Data Ocean into a Dashboards using huge data sets defines success.
Thank you for sparing time going through this article, kindly share your views and what you would like to see and hear from me in my future blogs.
Outstanding Outliers “AG”.