Hi ,
In 2004, Australian rower Sally Robbins made headlines by stopping rowing 400m before the finish line in the Women's Eight Olympic
rowing final.
The Australian team was in fifth position at the time but had started strong and were only 3 seconds behind the first-place team. Ultimately, though, the Australian team finished last.
It didn't matter that Robbins had already rowed more than 1km. Without the last 400m, she may as well have not bothered.
Here's the thing...
Data scientists love building models and are fantastic at doing so.
But as statistics, such as the much-quoted "87% of data science projects fail to make it into production" demonstrate, many data scientists fall down in the last leg of the race.
Popular model-building framework CRISP-DM lists the six phases of any modelling project as:
- Business Understanding
- Data Understanding
- Data Processing
- Modelling
- Evaluation
- Deployment
Quitting a project before the deployment phase is the equivalent of dropping your oar 400m
before the end of the race.
To win gold, you need to keep rowing to the end.
Talk again soon,
Dr Genevieve Hayes.
