Welcome to another episode of our ‘Machine Learning’ Series.
So is Machine learning the ultimate goal of any data science initiative? It seems to be so in the recent past where a quaint little area of Data Analysis is thrown into the cauldron of Machine Learning and AI madness. Folks (I admit including myself) got into to this area only due to the allure of Machine Learning and AI. It is like the bait that is catching and reeling us all in. I was listening to this podcast ‘path to a data science career’. In this podcast, they were talking about what does one needs to know to build a career in data science. The most important take away for me from this podcast is their discussion on if you need Machine Learning skills to be a good data scientist. The answer is NO. Or it is a qualified ‘it depends’.
Data Science is a vast interdisciplinary area. Some folks come to data science from a software programming background like me. Some can do the reverse and come from a math or statistical background. Or others from a technical field like biology or physics. So a team of Data Analysts will usually have a background from a diverse background that complements each other. So you can be wherever you want to be in the Data Science pipeline shown in the figure below. You could be great at Pandas package in Python and that may be all you do. If you can do it really well and build some great analysis packages for your colleagues to use. You could be a great report writer with skills in data visualization and storytelling. You can bring some technical background in understanding the data domain, say a geneticist or bioinformatician. If you are a machine learning engineer you can be building some interesting models on the data for better understanding, predicting, classifying or forecasting. Still you will only be part of this bigger team of experts and will depend on and complement their expertise.
No doubt with the newer ML algorithm and awesome processing power at our disposal, our ability to analyze and understand huge amounts of data (Big Data, another catchy bait reeling some of us in). Still most of the work of the data scientist/analyst is still being done in collecting, aggregating, filtering, transforming, hypothesizing and report writing aspects of Data Science. All these steps are critically important even if Machine Learning is the team’s ultimate goal. So pick a spot in this pipeline that interests you and where you can develop a deep understanding and experience. Don’t be a shallow grazer and a Machine Learning wanna be just because that is what the latest buzz is all about.
