Google Flu Botched Predictions, Overestimated Flu Amid Record Season
February 18, 2013 11:44 AM
Typically accurate algorithm may need tweaking to deal with new trends
Launched in 2008
Google Flu Trends
is one of Google Inc.'s (
) most intriguing internet services. Leveraging Google's mastery of data mining and industry-leading search engine position, the system injects data on health-related searches into a complex model which spits out an estimation of the number of people in each region that are infected with the flu virus.
I. Flu Trends' Bad Forecast
The system performs extraordinarily well, typically matching up closely with data from the
Centers for Disease Control
(CDC). The system typically is several days ahead of CDC predictions in 29 countries and today also tracks a second disease -- dengue.
But this flu season was
a rough one
for the Google service. Google badly over predicted the Christmas peak of flu season, estimating that over 10 percent of the U.S. population had flu, versus the CDC's data, which showed only around 6 percent had it.
Google Flu overpredicted, while Flu Near You underpredicted. [Image Source: Nature]
So what happened?
Experts believe part of the problem was public fear influencing search behavior in an unusual way. The flu season got off to its earliest seasonal start in nearly a decade, starting in November and peaking in December. Of the three major flu virus strains, the most virulent strain -- H3N2n -- dominated, as in the severe 2003 outbreak.
There were multiple deaths, particularly among the elderly from the outbreak. As a result, the mass media latched onto the story and focus a great deal of coverage on the flu -- more so than usual. Experts say that may have led people to increase searches for flu terms, even if they didn't have flu.
It was also a bad season for norovirus outbreaks, which may have led to people searching for flu terms. Norovirus is a separate pathogen that manifests itself via intestinal symptoms. However, many in the public mistake norovirus infections for flu.
II. Self-Tracking Efforts Also Miss
Professor John Brownstein
that the miss by Google shows the difficulty in adapting algorithms to new variables. He comments, "You need to be constantly adapting these models, they don’t work in a vacuum. You need to recalibrate them every year."
Professor Brownstein helps maintain "
Flu Near You
Boston Children's Hospital
project which tracks flu in the U.S. via volunteer self-reporting. The 2011 program tracks a sample set of 70,000 people and has 46,000 participants (who report on their own health and health of family members).
Tracking the flu-virus in real time is a daunting challenge. [Image Source: Navco]
To Google's credit, Flu Near You missed nearly as badly, underestimating the peak infection rate. It peaked around a little over 4 percent. It's unclear why that underestimation happened.
Digital tracking of influenza-like illness (ILI) -- symptoms such as fever and sinus issues -- first began in 1985 in France with the creation of the Sentinelles network. Today France's GrippeNet.fr continues that tradition. Similar to Flu Near You, GrippeNet uses self-reporting from volunteers and has 5,500 participants.
The CDC's tracking comes directly from a network of 2,700 health institutions, which cover 30 million patients.
III. What About Twitter Tracking?
The misses of Flu Near You and Google Flu Trends open the door to new analysis projects. A pair of new efforts --
-- look to track Twitter posts to determine flu rates. Johns Hopkins University Professor Michael Paul argues that Twitter monitoring may offer less noise than search term monitoring, while offering big sample sets than self-monitoring.
He comments, "I suspect that passive monitoring of social media will always yield more data than systems that rely on people to actively respond to surveys, like Flu Near You."
Some are turning to Twitter for flu tracking.
But some are skeptical of Twitter-mining.
, head of the CDC’s Influenza Surveillance and Outbreak Response Team, argues that Twitter crowdsourcing is less effect that search-term efforts, because Twitter's userbase tends to be younger, less reliable users. She comments, "The Twitter analyses have much less promise."
The 2012 miss was the
worst since 2009
for Google, which led to some major algorithmic adjustments. In 2009 Google under predicted the H1N1 swine flu outbreak. That mistake was the subject of a
paper entitled "
Assessing Google Flu Trends Performance in the United States during the 2009 Influenza Virus A (H1N1) Pandemic
"Well, there may be a reason why they call them 'Mac' trucks! Windows machines will not be trucks." -- Microsoft CEO Steve Ballmer
Google Helps to Track Swine Flu, Twitter Catches Criticism
April 28, 2009, 11:29 AM
Google Predicts the Flu
November 12, 2008, 10:10 AM
PIQ ROBOTTM reveals its new artificial intelligence software
November 29, 2016, 12:59 AM
One more time - Happy Thanksgiving to Everyone Around the World
November 24, 2016, 4:00 AM
Google’s Smart Contact Lens Project gets halted for 2016
November 20, 2016, 7:00 AM
Cell Research Study shows African Americans have greater immune response to infection
November 10, 2016, 1:00 AM
UTHealth Clinical Trial Shows Progress Using Stem Cells to Treat Traumatic Brain Injury
November 8, 2016, 1:00 AM
Uber Partners with Circulation to Pilot Program Connecting Transportation and Digital Health Care
November 6, 2016, 5:00 AM
Most Popular Articles
What Can You Do with Your New Echo Dot?
December 3, 2016, 5:00 AM
Google has developed Deep Learning Algorithm to detect Diabetic Eye Disease
December 4, 2016, 5:00 AM
Foscam R2 Home Security Camera System – High Quality High FHD Security Video Footage with No Monthly Fees
December 1, 2016, 2:00 AM
The iPlugmate is an Excellent iPhone Flash Drive at a Great Price. (Deal Expires in 3 Days)
November 23, 2016, 1:00 AM
Microsoft Surface – Which Surface is Right for You?
December 2, 2016, 5:00 AM
Latest Blog Posts
In The News
Dec 7, 2016, 5:00 AM
e Guide: Mobile Security for 2017
Dec 6, 2016, 5:00 AM
Apple Car is Not Dead
Dec 5, 2016, 1:00 AM
Dec 4, 2016, 5:00 AM
Dec 3, 2016, 5:00 AM
Dec 2, 2016, 5:00 AM
Surface Ergonomic Keyboard
Dec 1, 2016, 3:01 AM
Chapeconense plane crash: Football rallies around Brazilian Team
Nov 30, 2016, 1:00 AM
How to Extends Your iPhone’s Battery Life
Nov 29, 2016, 12:49 AM
Nov 28, 2016, 1:12 AM
News: Fidel Castro
Nov 27, 2016, 5:00 AM
Nov 26, 2016, 5:00 AM
Changes in Social status affect the way genes turn on and off within immune cells.
Nov 25, 2016, 5:12 AM
Austrian far–right hopeful Hofer may back EU vote.
Nov 24, 2016, 4:00 AM
Final Fantasy XV Leaked Before Nov 29 Launch Date
Nov 23, 2016, 1:00 AM
Nov 22, 2016, 2:26 AM
Nov 21, 2016, 1:00 AM
HTC Makes Big Moves in China
Nov 20, 2016, 2:00 AM
Do you know who is the number one company in the word?
Nov 19, 2016, 5:30 AM
Foldable Cardboard ”EcoHelmet” wins James Dyson Award’s Top Prize
Nov 18, 2016, 2:39 AM
Scientists Discover Roundest Object Ever Spotted in Universe
Nov 17, 2016, 1:00 AM
More Blog Posts
Copyright 2016 DailyTech LLC. -
Terms, Conditions & Privacy Information