Hi ,
When using data science to solve a business problem, it's easy to
become overwhelmed by the magnitude of the problem you're trying to solve.
And overwhelm is the most dangerous of all emotions, for solving problems. Because when you're overwhelmed, that's when you give up and walk away - leaving you no better off than before you started trying to solve the
problem.
In fact, you're frequently worse off, because you've now convinced yourself a solution isn't even possible.
But what if there was a way to avoid that feeling of overwhelm, in data science and in life, in general?
As it turns out, there is.
In his book, Tiny Habits, behavioural design expert Dr BJ Fogg argues that the secret to avoiding overwhelm, and changing your life, is to start small and simple - even if the change you make is so small it seems ridiculous.
Small changes are
easier and require less motivation than large changes, so are more likely to stick. And once a small change has become entrenched, it will inevitably snowball into something bigger.
So, for example, if you want to improve your fitness level, rather than setting the goal of going to the gym three
times a week, a better approach is to start by setting the goal of doing two push-ups after each time you use the bathroom.
According to Fogg's research, the tiny habits approach can completely transform your life. But it's also a fantastic approach to achieving results as a data scientist.
When designing a data science solution, rather than trying to build a complex model to solve all aspects of a business problem, under all possible circumstances, instead, start by building the simplest possible model to solve one aspect of the business problem under ideal
circumstances. And then iterate on that solution, making improvements with each iteration.
Granted, the initial model you build is likely to only be of limited value. But limited value is still some value. And you can always do better in the next iteration.
And this approach has also been proven to work. In fact, this is the approach to developing computer vision models, described to me by Dr Jeroen Vendrig in a recent episode of Value Driven Data Science. You can listen to the episode HERE.
From small things, big things grow - even data science solutions.
Talk again soon,
Dr Genevieve Hayes.
