What Data Science Has Taught Me
I eat data, and fight with code.
I always tell people that data is not the new oil, instead, it is the new time machine.
Today, using data, you can know how much profit your firm will attain in the next 5-10 years, and you can do a deep analysis of what could have been the reasons for the success and/or failures in the past decades or so. Observation is key here. A small change in your data can change the decisions your firm takes. That's a big challenge: to depend on data for running a business successfully. And that's a risk that has helped pile up profits for everyone. Think about it!
Here's what Data Science has taught me:
1. What You See And What Exists Is Different 🤯
Quite often, this might happen to you, while dealing with data, that the patterns you see with the naked eye, are almost always never, the patterns that actually exist. That's where the statistics come in. So many statisticians are moving towards data science, and it makes perfect sense because data science is basically based off of statistics. It's just that, it shows a much deeper relationship between the business ideas and how statistics affect them! 🤯
2. Data Science Is Not Data Analysis 🙄
Data Science is a vast ocean, out of which the analysis part of it, is concentrated in a small area only. Data Analysis pertains to checking the data for anomalies and building insights out of it. Data Science, on the other hand, corresponds to the questions which can be answered using that data and building an understanding of how the answers to these questions can help the business being analyzed.
Exactly, business analysis. It is the key to a data science role. It can be product analysis as well, but all in all, data science should answer the questions that have kept businesses in the dark. That's the difference.
3. It May Sound Weird, But Machine Learning Is A Pre-requisite For A Data Science Role 😵
Over the previous decades, tons of customer data has been collected, using methods such as Surveys. Today, in 2020, data is being collected from machines, to understand their behavior, in order to automate them for the coming future. 2020 is a perfect example of how the dynamics of the World will be changing in the coming years. Robotics is rising in the industry, day by day, and there are countries where robots are serving in restaurants! We sure have evolved as a species, and this innovation shows that.
In layman's language, machine learning is nothing but a study/understanding of how machines act and react to user inputs, as well as on their own. Data is essential for machine learning. For example, a robot's movements are captured, and then algorithms are run on that data, to predict how often the robot will function the way it is required to.
Understanding how the machine works, and what the outcomes of these algorithms represent, is also important, because these days, the role of a data scientist is very holistic, and deals with lots of cleaning, wrangling, and mining of data, but at the same time, it also deals with a deep understanding of machine learning algorithms that help us make the predictions (of a robot's expected working, for example).
4. Cloud Is A Powerhouse, And Data Demands It ✈️
Cloud technologies embed a stable and vast powerhouse in themselves, and that has become the sole reason for organizations to migrate towards the cloud, for storage of Big Data, and for running high-performance applications. Nowadays, people are working on GBs, TBs, and higher data storage, with tons and tons of data to analyze. On such extensive data, it is not easy to run algorithms on the whole data, in one go, on a normal PC. A Supercomputer is required for this task.
Supercomputers are not practical for organizations, and this becomes a strong reason to shift towards cloud technologies since they allow data to be stored, fetched and analyzed in packets while being added to the same algorithm, and the results being saved on the cloud as well. You could say that the whole infrastructure shifts to a cloud and that mind-blowing!
5. Why Data Is A Time Machine 👨💻 👩💻
Let's take the classic example of this year: 2020. Organizations are laying off people? Why? They sure haven't lost all their savings yet, and they sure are earning profits (reduced, of course), but organizations that have shifted online/work-from-home are still making some profits. So, why lay off so many people?
Because data helped them predict the coming recession, and they are planning for it. In a few days, you will see fruits and vegetables in your local market become costly. Gadgets' prices have already begun to soar. This is a result of the catastrophe that we are facing this year. So, data is helping companies make decisions.
And it has been happening so for about a decade now. This is why you see so many emerging startups that were built in the last ten years, and they have progressed with the speed of light! Because their decisions were built on a thorough analysis of incoming data, through years. The World is changing at a very fast pace, and data is determining these changes, bit by bit.
Data-driven decisions are the need of the hour, as we have seen how the World can change in a second, and we have experienced this, in 2020. Preparation for the future is essential, now that we understand how quickly our lives can be altered.
We can prepare for the future, in any domain, with the use of the existing data that has been collected over the years, and the data that is incoming (since the amount of data being collected each year, is also increasing at a rapid rate).
Thank you for reading! Happy learning!
After spending an enormous amount of time on the internet, and struggling for years, to keep my foot in the right direction, I have understood how helpful guidance can be! Hence, I have decided to contribute to society by keeping all my personal projects and learnings open-source.
I'd be really happy if you could support my writing by getting me some coffee! 😃
Subscribe to get your daily round-up of top tech stories!