Data Science is a hot topic at the moment and it is becoming even hotter now the large organisations and the bigger vendors like Microsoft have picked up on it as well. Many organisations are looking at it and want to hire a Data Scientist. However, they end up summing up requirements for, or even hiring, a Data/Reporting Analyst. This is causing some confusion and the fact that more and more Data Analysts are adopting the job title Data Scientist on their CVs isn’t very helpful either.
So how do we distinguish a Data Scientist?
I think firstly, it is good to understand what it is that a Data Scientist does.
A Data Scientist:
– Deals with data, meaning they collate, process and transform data so they can work with it;
– Through Exploratory Data Analysis aims to better understand the data’s meaning, it’s type, attributes and relationships between each other;
– Then works out statistical analysis and modelling to establish correlation hypothesis (X will influence Y, A and B are associated, C has an inverse influence on D, etc);
– Tests/validates the hypothesis by running statistical testing to conclude if the hypothesis is true or false;
– Lastly, they conclude their findings into insights and share this with the business.
And in modern days, Data Scientists also get involved in Machine Learning projects in which they construct algorithm models to:
– Predict odds in the future (linear regression); and/or
– Detect segments of customer behaviour (logistic regression, Tree, Forest algorithms); and/or
– Find customer’s preferences (Support Vector Machine); and/or
– Build recommendation engines (Collaborative Filtering and Support Vector Machine); and/or
– Set up fraud and risk analysis, and much more (Principle Component Analysis).
Common analytic tools used for these activities are SAS, R, Python, Matlab and SPSS. Although the latter is becoming less influential in the modern field of Big Data/Data Science.
The keywords you should be looking for in a true Data Scientist are “statistics” and “algorithms” and find out what models they have set up to distinguish them from traditional Data Analysts. These backgrounds and experiences are typically lacking with Data Analysts. That is not to say that a Data Analyst can’t become a Data Scientist! With the right knowledge and application of statistics and algorithms, this is definitely possible.
I also like to add to not take a job title that a hiring company puts on one of their vacancies for face value. Data Science is a new area for most organisations and not everyone understands the concept yet. So ask questions about what it is they need to get done and what technologies they are adopting. Are they truly looking for a Data Scientist?
** A special Thank you to Jixin Jia – Microsoft BI & Advanced Analytics Consultant and public speaker – for your time and input to help me understand the world of Data Science **
If this blog was useful to you or you would like to comment, please feel free. I welcome discussions and feedback.
If you want to have a more confidential dialogue, feel free to contact me at firstname.lastname@example.org or 0385-927-827. Also, feel free to check out my bio page on our website. There you’ll find my latest jobs, social media feeds, and a little more about me.