Share with your network!

Have you ever wondered how Youtube magically lists up the videos of your taste in its suggestions?

Well, that’s an outcome of Data Analytics. The word data analytics has multiple meanings. It depends on the context of usage. The general understanding of the term is that when we obtain a set of data points from the subject under test, we try and make some sense of that data. And we try to understand the subject better.

For example, the Youtube collects the records of the previous videos you have watched and tries to find a video which is close to the set you have watched. Sounds simple?

There are three main steps which go behind the process of data analytics. First is the “Problem definition”. It’s very important to define specifically what you want to know from the data before doing any analysis. This will be the base over which you will use other techniques to find the required relation or trend from the data.

Then, the second step involves applying statistical techniques to process the data. One example is to predictively evaluate what a user likes to watch by finding correlations between data using statistical tools. This is part of predictive modelling.

Statistical analysis has several sub steps:-

  1. Cleaning the data: In this step, gross errors present in the data are addressed. Say your address is from a form filled by patients. Spell checking would be a cleaning process which would ensure that wrong spellings wouldn’t harm your analysis process.
  2. Assessing the quality of the data: It is very important that we know how good is our data before we try to make some analysis on it. Mean, standard deviation measurement, etc are few of the basic methods used to find how good the data is. The exact tool which is used would depend on the analysis being performed.
  3. Main Data Analysis: The main data analysis could be of two types, exploratory or confirmatory. Exploratory analysis is done to find any new relationship in the data. While confirmatory analysis is done to strengthen or weaken the already existing relationship in the data.

Once the above three steps are done, we are finally ready to answer the main question we have set forth with. The third step is to answer the question by using the relationship we have deduced in the data. Generally, computational programs are written to do this step.

Data analytics goes by various names in various fields. Social network analysis, financial engineering, Business Analytics, computational neurosciences, etc are few of such names. The philosophy behind all these fields is same. They try and understand data collected from real life situations and make a link out of it and try to apply to a given problem and solve it.

One interesting example is cognitive neurosciences. It’s a fact that we don’t understand brain completely. But, this field tries to find patterns in our brain’s behaviour and hence deduce what a person is trying to think by finding patterns in their nerve impulse triggering. Imagine an app that determines what a person is thinking based on his brain wave reading (that’s gonna happen soon apparently, we will write in future about it. Subscribe using the bar to read about it when published). This example shows how powerful Data Analytics is.

We can without fully understanding a subject, try and deduce useful information from data. Data analytics is all about doing this. It’s a very fascinating field. Hope this article triggered enough interest in you to read further about Data Analytics! Stay tuned by subscribing to our newsletter to get the latest updates now!