Mayer-Schenenberger and Cukier have pointed out that
“big data” is about PREDICTIONS. It applies computational
algorithms and mathematics to huge quantities of data,
often, “messy data,” and infers predictions.
This active trend affects traditional sampling plans and
sample distributions since, at its aim, it collects all the data
and can, as a result, provide a clearer view of the “granularity
of the data”– the sub-categories that smaller samples can
Further predictive analytics, based on correlations, detects
directions and inferences, yet does not seek causes or
test hypotheses, as it alerts us to what is happening.
-mechanical or structural failure predictions based on heat,
vibrational, stress and sound patterns from sensors,
-hit songs and TV programs
-data mining drug candidates
Counter to our intuition where we evoke “causality,” in which case,
as Kahneman says, our brain is too lazy to think slowly, we jump
to shortcuts. Big data analytics provides a “reality check.”
Take the case of Louis Pasteur “curing” rabies in the nine year old
boy, Joseph Meister, by inoculation in 1865. Looking at the
data, on average only one in seven people bitten by rabid dogs
ever contracts rabies. (85% chance he would survive without
ACS needs to serve its members by continuing to collect data,
but broaden its outlook on how the data can be “mined.”
Mayer-Schonenberger and Cukier document that data is
(1) reused, after first use, (2)merged with other datasets to explore
new venues and (3)”extended”. By extended, we use the “data exhaust.”
There is so much more ACS can do to serve its members,
just being open to new thoughts and emerging trends and not
feeling we have done it before or falling to the NiH syndrome.
(NiH = not invented here)
Where are chemical enterprise careers moving? What skills
will be needed? PREDICTIVE ANALYTICS
How do we effectively help members gain advantageous skills?
What knowledge, approaches, methods and skills should be
offered to professionals? How can it be done cost effectively?