29 September 2020

Event Driven Data - part the second

Yet another serial. Events drive data, full stop. Most of the time, the data science cabal acts as if it's the other way around. In the realm of God's Rules, where the rules don't change (and we puny humans know them all; see long term weather forecasting with massive supercomputers), data, in particular time series, are the product of a stable process. 

Human driven processes aren't like that. The Super humans get to change the rules as they see profitable. And sometimes, not nearly as often, it's the Little People that drive the structural change. Covid-19 is one such instance. 

Now comes this analysis, courtesy of McKinsey, the Stasi of the management consulting sector.

But the [Covid-19] crisis has also revealed the technology's Achilles' heel. One of the most widely used advanced-analytics techniques—machine learning—relies on the principle that patterns and behaviors from the past will likely repeat in the future. [my emphasis]
The point is, war-gaming is what's needed to recalibrate
While analytics professionals will play an important role, successfully stabilizing critical models will depend equally on efforts from leadership to recalibrate business strategies for the changing landscape, forge new data partnerships, convene interdisciplinary teams with sufficient diversity, and more.
Machine learning is nice. Reading the NYT every day is too.

No comments: