Creative Mindset - The Unsung Hero in Data Science

In the world of Data Science, relying solely on theory is like appreciating a song just for its lyrics. To truly comprehend the song of numbers, one needs an artist’s touch.?It's only after mastering the rhythm of data that one can tune up the scientific toolbox and crank up the symphony.

The artistic perspective isn't confined to intuition building. This mindset is pivotal at every juncture, from collecting data to selecting features and modeling. The ability to identify subtle nuances and connections within data sets apart great data science practices from good ones. This article delves into the interplay of these elements.

The Old days

In an era devoid of digital footprints, the art of observation was as valuable as data itself. In the shopping malls, the quiet observers would glide through, treating the aisles as their stage and the shoppers as their subjects.?

Ah, the joy of deciphering customer behaviours without SQL queries!

In his book 'Why We Buy,' Paco Underhill mentions that artists were the primary interpreters of customer behavior, followed by experts refining insights. Artists delved deep into the human mind, not just noting a shopper's cursory look but grasping the emotion, the hesitation, the internal debate. Their innate intuition and ability to embrace uncertainty allowed them to see patterns often overlooked by more methodical scientific approaches.

Creative minds, free from rigid theories, observe patterns with innate curiosity, laying the groundwork for retail research in the pre-digital age.

Their insights influenced store designs, window displays, and the arrangement of aisles. Concepts like the "butt-brush" effect were brought to the forefront, highlighting the subtle nuances of human behavior.

Evolution of Data Science

Back to the new age—a time where your coffee machine talks to you in the morning (hopefully, not about existential crises)

Data Science's rise to tech stardom wasn't overnight; it evolved, much like a butterfly from its cocoon. This innovation? The blend of artistry with algorithms.?Because while algorithms are advanced, without personalized connection, they're as practical as a hammer in catching fish. It's the artistic mindset that lets us think outside the Cartesian plane.

Let me share a few brief stories where you will learn how creative edge is not just a decorative frosting; it’s a key ingredient. Omit it, and you risk turning your Data Science masterpiece into, well, a science experiment gone wrong.?

Siri vs Cortana

Apple's darling, Siri, wasn't always the quick-witted voice assistant we adore (or occasionally argue with) today. In its inception, Siri was like a toddler—filled with potential, but could barely tell a cat from a hat. It took artists, storytellers, and humorists combined with engineers to give Siri her personality. The magic? They realized she wasn't just a tool, but a companion. Making Siri feel 'real' was all about crafting those quirky responses and ensuring she understood the emotional undertones of our questions. So, the next time she gives you a sassy response, remember she's been to art school.

Microsoft’s Cortana tried to sing in the same choir as Siri but ended up hitting more sour notes than a cat walking on a piano. Why? Well, while Siri was attending art school, Cortana was all about the science and none of the art. She was engineered to be a hyper-efficient, super-logical personal assistant. And that’s where she failed. No sass, no warmth, no soul. It's like dating someone who knows all the statistics about love but forgets the first rule—actually falling in love.

Spotify's Symphony vs Pandora's Missed Notes

Spotify's Discover Weekly is like having a mixtape from a friend who just gets your music taste. While algorithms did the heavy lifting, it was the artistry that fine-tuned the experience. Instead of simply focusing on genres or beats, the artists (data scientists) behind the scenes considered the emotions, narrative arcs, and even the cultural significance of tracks. The result? Personalized playlists that felt handpicked by a BFF who’d raided your diary. And maybe, just maybe, cried at the same songs as you.

Pandora's Thumbprint Radio was an ambitious project. They had the algorithms, the data, even the funding. What they lacked was the 'creative edge' that Spotify brought. While Discover Weekly is like your cool friend, Thumbprint Radio was like an overeager parent trying to set you up on a date. The recommendations were based more on mathematical patterns rather than the emotional resonance of the songs. It's the difference between a mixtape and an Excel spreadsheet—both can list songs, but only one makes you feel.

The Netflix Story

The story of Netflix, as it transitioned from a 'DVD by mail' service to a streaming behemoth, offers an emblematic example of the synthesis of art and science. But did you know that one of its groundbreaking features was influenced by wartime tactics?

During World War II, statisticians were trying to determine where to place armor on fighter planes. By examining where bullet holes were found on returning aircraft, one might assume the solution would be to place armor where most holes were found. However, a clever statistician named Abraham Wald realized the opposite. The planes were returning despite being hit in those places. It was the areas without bullet holes on returning planes that needed armor since planes hit there weren't making it back. This counterintuitive approach is an early example of latent feature recognition.

Fast forward to the Netflix era. To perfect their recommendation engine, Netflix didn't just examine what movies people were watching. They dived deeper, identifying latent features, characteristics that aren't directly observed but inferred from data. Similar to Wald's planes, Netflix identified viewing patterns and preferences that weren't explicitly stated by the users but were critical for tailoring recommendations - a founding stone for 'Similar User Behaviour' recommendations.

Conclusion

Step into the world of our exemplar creative Data Scientists - they might borrow a note or two from Spotify or so, but they orchestrate their own symphony. After all, copying is not art!

While Art ignites the spark of imagination and sees potential in the chaos of data, science provides the blueprint and execution strategy to bring that vision to life.

But what happens when you blend these worlds? - The magic begins. So how do you get started?

Harnessing the Art in Data Science

So, how can you keep up with creative edge? It's about telling a compelling story with that data.

Here's how to infuse your data science with an artist's touch:

1. Brainstorming Sessions: Organize sessions with diverse teams, encouraging unconventional ideas. Visual aids can help map out connections.
2. Diversified Learning: Consider taking short courses in storytelling, design thinking, or even fine arts. This can enhance your ability to represent data in a manner that's both informative and engaging.
3. Data Visualization Workshops: A well-crafted graph can convey more than a thousand numbers.
4. Feedback Mechanisms: Establishing a continuous feedback loop with stakeholders ensures that your data output doesn't just make logical sense, but also resonates with the target audience.
5. Case Study Analysis: Review past projects as well as look outward to refine strategies. Encourage discussions to delve deeper into decision-making.

In the time it took to read this article, countless data points were generated, and somewhere, a LinkedIn Data Scientist [uhmm... artist] looked at this article and had an "Aha!" moment.