The Power to Transform Big Data into Knowledge | Montserrat Fuentes | TEDxRVAWomen


Translator: Ivana Krivokuća
Reviewer: Denise RQ I’m so excited to be here
with all of you today. I grew up by the clean beaches
and the beautiful, green mountains of the northern part of Spain. It’s just relaxing and inspiring
looking at this beautiful picture of my hometown, isn’t it? I didn’t want to ignore
that beauty of nature that I missed so much when I moved away. I also noticed the levels of stress
started increasing as I would spend time away from nature. I didn’t want to become
made out of concrete that I started to see
more and more around me. Along with nature, music was always
a great passion for me. Playing the piano
would relax me, inspire me, and also, was a way for me
to express myself. I also liked math, numbers.
I liked the logic in math. Math for me has been always a way
to make sense of the world around me. Math and music could be perceived
as a very odd couple. But for me, they both define me.
So I study both. Actually, there is a lot of math in music. Musicians need to understand
fractions very well if they want to hold correctly
the value of the notes. For me, music would enhance
my math learning. The answers to a lot of math problems would occur to me
when I was playing the piano, so then I started to put
my math homework on top of the piano. It’s what I hoped, as I was playing, that I’ll be able to figure out
the answers to my homework. I wanted to use numbers. I wanted to use math to be able to address
some important real world problems; problems that mattered to me –
environmental problems. Then is when I learned about statistics;
that is a branch of mathematics that helps you to transform
numbers into knowledge. Statistics is the science
that extracts information from data, and we apply that
to the environment or to medicine. We have introduced personalized medicine. What that means is that we developed
individualized treatments for patients, and that helped us to improve human health by combining a lot of data from genetics
and the environmental conditions. That’s the beauty of statistics. Statistics helps you to use
your technical knowledge and skills and apply it to any area
that you are interested in. Statisticians often say that a great thing
about being a statistician is that statisticians can play
in everyone’s backyard. And I love backyards, particularly the ones that have
a lot of green areas and beautiful trees, so I can bring my children
in the beauty of nature. But then, soon enough, I learned
some of the challenges that women in science often face, and I was determined to overcome them. For my daughter, I have always wanted
to be a good role model for her. I was told I couldn’t lead scientists
because the important conversations, the important decisions
at the universities happen in the men’s bathroom. (Laughter) Well, I led scientists, and now at VCU,
where I’m very proud to be, I lead the humanities and the sciences, and I never, not even once,
had to use the men’s bathroom. (Laughter) (Applause) One of the most important trends
right now in our society, probably the most important
trend in science is the access that we have to big data. We have big data everywhere. So it’s important
that we make use of the data, that we transform the data
into valuable knowledge. Let me just illustrate
what really big data is about. We have now the Square Kilometre Array. That’s the world’s most powerful
radio telescope. This telescope in one day is able to generate as much data
as we have on the Internet today. It has enough optical fiber
to wrap twice around the Earth. It has the processing power
of one billion PCs, and in one day generates enough data
to fill 15 million iPods. That’s big data. We want to make use of the data, we want to introduce the statistical tools
that can help us to transform the data, give us information, knowledge. We need to develop those approaches
that can help us to work with big data, that we can summarize the results
and be able to reach the public with powerful visualization tools. And it’s just much more than the mean. I’m sure you know
the statistical mean, the average, but it is much more than that. Let me illustrate some of the limitations
of working just with the mean. Here we have 100 pictures of kids
going to see Santa. There are some questions of interest. We want to know if we have maybe more boys
or more girls going to see Santa, if they tend to sit
on his right lap or the left lap, or we get two kids
at the same time versus one. Those are the questions of interest. Well, let’s take the mean
to see what happens. What does the mean? That’s what we get.
You see it in the picture. It’s not even clear that’s Santa.
That could be the wizard of Harry Potter. (Laughter) So we need more than the mean, we need to be able to summarize
the data in different ways. We need to look at percentiles, and we certainly need
to look at the variability that we see across the pictures. We need to characterize the uncertainty
in our mean, in our estimate. Statistics is called
the science of the uncertainty. When we provide
prediction, forecast, a mean, we need to be able
to characterize the uncertainty that we have about that estimation. I love uncertainty.
It makes me comfortable. I’d rather go through life
with acceptance of not knowing, than just having an answer
that is simply yes or no. In my research, what I do,
I introduce the statistical methods, the statistical tools
that can help us to address some of the most relevant
environmental problems: how the temperatures are increasing,
how the level of pollution is increasing, and how is that impacting human health, leading to premature death
or birth defects, when we have pregnant women
exposed to those high levels of pollution. As I left the northern part of the Spain,
I developed a passion for the environment. I wanted to bring the calm,
the beauty of nature to my children, and I use statistics to address
those environmental problems. In 2003, we had in Europe
35,000 people dying due to exposure to very high temperatures. So with my statistical methods,
I look at the temperatures – how they are increasing over time. Here I’m projecting,
going from 1990 to 2040, in the southern part of the United States, we are going to experience an increase
of 5.2 degrees Fahrenheit in temperatures. If we look at
the very extreme temperatures, the very high temperatures,
temperatures above the 95th percentile, the increase will be
7.6 degrees Fahrenheit. That’s a very significant increase. I also look at the number of days
exceeding a very high temperature. For instance, 99 degrees Fahrenheit. When we move into 2040,
I’m projecting it’s going to be 54. I also characterize the uncertainty
about that prediction. So it will be between 46 days and 64 days. That’s a very, very significant increase
in the temperatures. The increase in the temperatures is also going to be linked
to increase in air pollution, because the air pollutants,
for the most part, are very sensitive
to the temperature values. In particular, the ozone. The ozone is a very dangerous pollutant. It leads to cardiovascular problems,
premature death. For the formation of the ozone
we need high temperatures. So we project, and now moving
from 1990 to 2050, we are going to also observe an increase
in the values of the ozone. That increase is going to change
across the space. It’s going to be greater in the Midwest. We are going to experience
about 12% increase in the Midwest. Along the coastline,
the increase will be about 4%. So we look how the pollution
is impacting human health. We introduce statistical models
that help us to link the temperature, air pollution, the human health outcomes, with the purpose of helping,
trying to help our citizens, to improve the well-being of our citizens. But we need to summarize the results
in a way that can reach you, so we use maps, we use colors. In this map, when the color is green, that indicates that we have
acceptable levels of pollution. As the color goes from green
all the way to purple, that indicates an increase
in the levels of pollution. We also have flags in the schools. In North Carolina, we have a flag to indicate
what the levels of pollution are. When it’s yellow,
kids with asthma should be inside. When it becomes orange, we keep
all the children indoors during recess to avoid exposure
to very high levels of pollution that could lead to respiratory,
cardiovascular problems or asthma. We have such a great demand
for statistics, and a small supply of statisticians. It’s just a great, great time
to be a statistician. (Laughter) Actually, statistics was named
the sexiest job of this decade. (Laughter) (Applause) Statistics helps us
to transform data into knowledge, protect our environment,
make informed decisions, make the world a better place. And remember, most likely,
you have a statistician in your backyard. Thank you. (Applause)

Leave a Reply