It’s placement time across India. Amidst times of demonetisation and the Trump regime, Rajesh decides to go for a job and postpones his plans for a master’s for another ‘x’ years. Here, the value of ‘x’ is dependent on the probation period. But, yes, Rajesh previously wanted to join core companies for a job offer but now he prefers the job of an analyst because of two reasons: lucrative offers and since it’s more popular these days. The analytical abilities required for being a data analyst are possessed by all up to a certain extent. But, yes, Rajesh believes he has an edge as he has published a paper in statistics.
The above scenario is one of the most common phenomena happening these days. The number of people enrolling for online courses on data analytics has increased exponentially. From what was once considered as a branch of computer science engineering, data analytics now finds its place even in the areas of electrical engineering, mechanical engineering. So let’s first understand each fragment of this area one by one.
Data is a piece of information to be put into simple words. It can be numerical, example — age, weight, cost of an item, etc. or even categorical like skin colour, the flavour of cheesecakes. So what’s the use of data? Well, data needs to be processed just like the food you eat. After the processing, we can have some results which can be used for deducing suitable inferences. Sounds like those lab experiments! Let me explain to you with an example — there’s a famous restaurant opposite VIT University, Tom’s Diner. They make the most amazing cheesecakes in Vellore. So what do I do? I collect the sales record of the different flavours of cheesecake sold in a week.
Now, I process it using simple statistics that we all know — bar graphs. The results clearly depicted the sales by flavour. So what are our conclusions? The most popular flavours and the total number of units sold per week. We can suggest that the chef should make more of these flavours as they’re popular.
So yes, whatever we do after the collection of data is called ‘analysis of data’. Our neighbours, relatives are actually more popular for doing the analysis of our data — Facebook check-ins and WhatsApp stories, etc. But yes, like the above example, data analysis can be biased too. Surprising, right? Like supposedly, your relative sees a glass of Appy Fizz in your glass and he assumes it to be alcohol and says he saw it on your wall for say, around five times. There you go, officially a drunkard pronounced by your relatives. But say for the next 10 posts it never happens, so yes you may have drunk occasionally say in an office party, but still, people assume you to be drinking like every time. So that’s a bias created (weird but technically the same word is used to explain this abnormality).
There is also a need to understand the importance of making sense of the data. Biases are possible to make favourable views, but we need to eliminate bias for a successful analysis for most cases. Example — while planning a city, we should primarily take in views from educated citizens and city developers. Here, we created a bias to simplify our understanding but this may not be the case in all the instances, example — measuring awareness about diseases and their spread. Bias and its need depend on the problem.
By now you must’ve got a clear understanding of data and analysis of data. So what does a data analyst do? Well, like our relative, their job is to analyse the data and suggest some possible inferences. This does not imply that statistics alone is used, there are other optimisation methods like genetic algorithms, neural networks too. Statistics essentially form the foundation of analysis but like Rajesh, we cannot say that statistics alone can help us land a job as a data analyst. We need analytical skills with the ability to correlate to certain areas or problems or practical applications.
But, yes, we do have a large volume of data available at all times. Yet, it is never analysed. For example, we all want to become data analysts but very few understand what they want to analyse. Data analytics can help in making better decisions but yes we need to make sense out of the data. The average rainfall can help us design better water harvesting systems in a city. The money spending break-up can make us visualise where we need to cut down our budget. There is a strong need to make sense of the data. There can be an issue regarding verification of data, that’s why we need a unique verification method, like a digital fingerprint to verify the data at all times.
Probably this article would make some sense in understanding the essence of data analytics and how it works. More importantly, simplify the understanding the foundation of Industry 4.0. In my next article, I’ll explain Industry 4.0 and its implications or as I say decode it for simple minds. So, stay tuned and keep making sense of the data.