Welcome to Decision Data
There are many decisions that we have to make when dealing with data: choosing the right chart, selecting algorithms, deciding between evaluation metrics — the decisions are endless, and it can be challenging to know where to begin.
When I was completing a subject in data mining for my undergraduate degree, our lecturer told us, “If you take the best data scientists from around the globe and present them with the same problem, you will find that no two results will be the same.” Considering I was studying mathematics at the time, this statement was discomfiting to me. The thought of transitioning from the study of a (generally) consistent system of thinking where results are derived from established, rigorous and widely-accepted perceptions of the universe to a career thriving off rapid innovation and disruptive action with concepts amalgamated from a variety of fields — let’s say that this frightened me enough to swiftly enrol in a series of weekend courses for SQL databases.
If you take the best data scientists from around the globe and present them with the same problem, you will find that no two results will be exactly the same.
It was only two years after I graduated that I noticed an emergence of formal data science degrees. By then, I had already tutored a data mining class, completed two data analyst internships, and began full-time data analyst work. There is a part of me that wishes these data science degrees had emerged much sooner. Still, that early period of observation, exploration, perspiration and making juvenile mistakes (which I still make) was crucial to the development of my intuition.
Yes, “intuition” — that confusing concept describing how we approach and make decisions from our data. Developed individually through years of trial and error and reinforced by our various backgrounds and experiences, it seems that what we perceive to be “common sense” may not be that common.
Whilst there might be a plethora of ingestible materials out there, I am yet to find one dedicated to reducing decision fatigue — especially for all the budding data scientists whose first instincts are to “try everything and see what works”.
In my journey, I have learned that I make better decisions about which path is appropriate when I understand the properties of the data space, as it echoes the rulesets behind mathematical and statistical concepts. This blog aims to share what I have learnt in establishing my decision-making processes with all developing analysts.
Welcome to Decision Data.
