This document provides an overview of how data science can benefit businesses and examples of how data science has been applied in different industries. It discusses how data scientists can help businesses harness big data by performing tasks like customer segmentation, predictive analytics, forecasting, and developing data products. The document also provides several case studies and examples of how specific companies have used data science for applications such as optimizing marketing strategies, reducing customer churn, and improving supply chain management.
3. I did Data in ShanghaiI did Data in Shanghai
Helpinganeventwebsitetargetitscustomers
4. Data in LondonData in London
Datageeksneedtoolslikethese,Ihelpedlaunchthebeta.
5. I worked with Data in LuxembourgI worked with Data in Luxembourg
Forasmalle-commercewebsite.DoingSupplyChainmodels.
6. Now I work withNow I work with
Data in Luxembourg @Data in Luxembourg @
7. In Air Traffic ManagementIn Air Traffic Management
BTWAirtravelproducesALOTofdata!
8. So how come I care about data?So how come I care about data?
Well I always loved science.Well I always loved science.
I wanted to be a neuroscientistI wanted to be a neuroscientist
9. Then I fell in love with PhysicsThen I fell in love with Physics
I studied Quantum Mechanics andI studied Quantum Mechanics and
Quantum Optics at BristolQuantum Optics at Bristol
10. This gets a bit complicated...This gets a bit complicated...
And my cat was never much use...And my cat was never much use...
11. So I ended up in Math & Stats...So I ended up in Math & Stats...
12. Along the way I learned someAlong the way I learned some
programming and other skills...programming and other skills...
13. I needed to find a careerI needed to find a career
And I decided Academia wasn't for me.And I decided Academia wasn't for me.
So I became a data scientist!So I became a data scientist!
Now what skills does a data scientist have?Now what skills does a data scientist have?
14.
15. To me data Science is a lot like ScienceTo me data Science is a lot like Science
16. But the hardest thing to learn has been..But the hardest thing to learn has been..
“Beingadatascientistisnotonly
aboutdatacrunching.It’sabout
understandingthebusiness
challenge,creatingsomevaluable
actionableinsightstothedata,
andcommunicatingtheirfindings
tothebusiness.”
Jean-PaulIsson,Monster
Worldwide,Inc.
17. I'm still learning the tech stuff tooI'm still learning the tech stuff too
What is Machine learning?What is Machine learning?
Well this next slide might help...Well this next slide might help...
18.
19. What is Data Science?What is Data Science?
Data Scientists help you harness theData Scientists help you harness the
value of 'big data'value of 'big data'
20. So, who is talking about Big Data?So, who is talking about Big Data?
"Weprojectaneedfor1.5millionadditional
managersandanalystsintheUnitedStateswho
canasktherightquestionsandconsumethe
resultsoftheanalysisofBigDataeffectively."-
,McKinseyreport
Bigdata:Thenextfrontierforinnovation,
competition,andproductivity
24. But talk is cheap...But talk is cheap...
Bigdataisliketeenagesex:everyonetalksabout
it,nobodyreallyknowshowtodoit,everyone
thinkseveryoneelseisdoingit,soeveryone
claimstheyaredoingit...-
ProfessorDanAriely-DukeUniversity
25. So how do you get value out of yourSo how do you get value out of your
data?data?
26. You could hire a data scientistYou could hire a data scientist
This talk is aimed at helping you understand if youThis talk is aimed at helping you understand if you
need to hire a 'data scientist'.need to hire a 'data scientist'.
28. Or this talk could have been titled...Or this talk could have been titled...
But Peadar where are the businessBut Peadar where are the business
examples?examples?
31. But I work in the real world not online!But I work in the real world not online!
UPSusesdatatotravelmoreefficientlyand
savemillionsonfuelconsumption.
32. What is "Big Data" anyway?What is "Big Data" anyway?
Customerpeferencedatais...
33. What is "Big Data" anyway?What is "Big Data" anyway?
Wind sensor data is tooWind sensor data is too
'Source:EnglishWikipedia,originalupload15July2004by
-CreativeCommonsSharealikelicenseLeonardG.
34. What is "Big Data" anyway?What is "Big Data" anyway?
Webcrawlingdataistoo
35. What is "Big Data" anyway?What is "Big Data" anyway?
Audioandvisualdataistoo
36. What is "Big Data" anyway?What is "Big Data" anyway?
SocialMediadatamustbetoo
37. What is "Big Data" anyway?What is "Big Data" anyway?
Nottomentionsocialmediametadata
43. Pick the right methodology for the jobPick the right methodology for the job
Text->topicmodelling,sentimentanalysis,information
extraction
E-commercedata->prospensityanalysis,collaborative
filtering
Multimedia->speech-to-text,audiofingerprinting,face
recognition
Clickstreamlogs->frequentpatternmining,sequenceanalysis
Proton-protoncollisionfromLHC->Ihavenoideadespite
havingaPhysicsdegree
46. What is a data product?What is a data product?
Adataproductprovidesactionableinformationwithoutexposing
decisionmakerstotheunderlyingdataoranalytics.
Examplesinclude:MovieRecommendations,Weather
Forecasts,StockMarketPredictions,ProductionProcess
Improvements,HealthDiagnosis,FluTrendPredictions,Targeted
Advertising.
–MarkHerman,etal.,FieldGuidetoDataScience
54. Case Study - Marketing Analytics: InCase Study - Marketing Analytics: In
the Game Industrythe Game Industry
1)Usesgamersplaydatatooptimizemarketing
communicationsacrosschannels.-Customersegmentation
modelling
2)BuildingPersonalizationEngineRulesfor1:1
communicationswithindividualgamers.Tohelpreducecustomer
churn.
3)Predictsgamerslikelihoodtochurnortorespondto
up-selloffers.
56. What about forecasting? Do you mean the weather?What about forecasting? Do you mean the weather?
(In Ireland and the UK it is quite easy - just guess rain(In Ireland and the UK it is quite easy - just guess rain
all the time!)all the time!)
But there are other kinds of forecasts such as supplyBut there are other kinds of forecasts such as supply
chain forecasts or demand forecasts....chain forecasts or demand forecasts....
58. But predicting the future is hardBut predicting the future is hard
"It’s Difficult to"It’s Difficult to
Make Predictions,Make Predictions,
Especially AboutEspecially About
the Future" -the Future" -
Niels BohrNiels Bohr
60. User Conversion after a website change.User Conversion after a website change.
Luckily they measured it. They learned the websiteLuckily they measured it. They learned the website
change was a bad decision.change was a bad decision.
61. Sometimes in life you just need picturesSometimes in life you just need pictures
or data visualizations, like this...or data visualizations, like this...
62. Or this: Number of wind turbines by state in the US?Or this: Number of wind turbines by state in the US?
63. But this is Luxembourg...But this is Luxembourg...
So we need finance examples...So we need finance examples...
67. But I don't work for a corporationBut I don't work for a corporation
DatacanbeusedforNGOstoo...
68. Data can be used for NGO'sData can be used for NGO's
Thisisawebappofhousepricesandcommutes,doneforanNGO
inLondonwhowantedtoshowtheeffectsof
changesinhousepricesonpeoplescommutes.
69. Reducing Maternal Mortality Rates in MexicoReducing Maternal Mortality Rates in Mexico
--Mexico - Presidencia de la RepublicaMexico - Presidencia de la Republica
ThematernaldeathsinMexicofrompregnancy,childbirthor
postpartumcomplicationshavedecreasedfrom89deathsper
100,000livebirthsin1990to43in2011.Despitethis
improvement,therateofdeclinehassignificantlyslowedand
MexicoisnotontracktoachieveitsMillenniumDevelopment
Goalofreducingmaternalmortality75%by2015.
70. But what if you're in Politics?But what if you're in Politics?
71. Earlier I showed how data canEarlier I showed how data can
even predict 49 out of 50 stateseven predict 49 out of 50 states
in the last American election.in the last American election.
AndthatObamawouldbethepresident!
72. So why would I need a data scientist?So why would I need a data scientist?
Youmayalreadyhaveone.Iknownumerousbusiness
intelligence,dataanalysts,businessanalysts,riskanalystswho
AREdatascientists.
Alternativelyyoucanhireadataanalyticsconsultanttohelpyou
getstarted.
ButwhatsignalsshouldIlookfor?
Welltherearemanyanswers...Like...
73. Does this sound like you?Does this sound like you?
Are you not taking full advantage of your reporting?Are you not taking full advantage of your reporting?
Do you need a high level visual overview of yourDo you need a high level visual overview of your
operations?operations?
Are you targeting your marketing efforts effectively byAre you targeting your marketing efforts effectively by
using the right customer segmentation - by age orusing the right customer segmentation - by age or
gender for example?gender for example?
74. Or this?Or this?
Are you losing customers and not understandingAre you losing customers and not understanding
why?why?
Are you making decisions on the basis of data or onAre you making decisions on the basis of data or on
the basis of 'gut feeling'?the basis of 'gut feeling'?
Are you changing your websites or products on theAre you changing your websites or products on the
basis of data driven experimentation?basis of data driven experimentation?
76. What about communication?What about communication?
Mathematicallysoundcommunicationtoclients:youmayhave
situationswhereyouneedthedatascientiststotalkdirectlyto
clientsortotheirdatascientists.
Thisisyetanotherreasontomakesureyouhiresomeonewith
excellentcommunicationskills,becausetheywillberepresenting
yourbusinesstoreallysmartpeople.
77. Data Scientists are like 'translators'Data Scientists are like 'translators'
Alotofmyworkatthemomentismathematicalcommunication
withexternalstakeholdersandProfessors.
AtAmazonalotofmyworkwaswithResearchScientistsin
Optimization.Translatingtheirideasforbusinessstakeholders.
Ioftenhavetotranslatefromthe'business'tothe'software'
team.
DoyouhavesomeonelikethatonYOURteam?
78. If this sounds familiarIf this sounds familiar
then you might needthen you might need
to hire a data scientist!to hire a data scientist!
79. I hope the examples helpedI hope the examples helped
Ihopeitisalsoclearhowdatacanbeusedinwhateverfieldyou
workin.
Ihopeitisalsoclearthat'bigdata'isnotsomethingtobescared
ofbutshouldbepartofyourorganizationsstrategy.
Iknowthatdevelopingadata-drivencultureisextremelydifficult.
80. Thank You For ListeningThank You For Listening
Anyquestions?
Reachouttomeifyouhaveanydataquestions.
@springcoil
SearchPeadarCoyleonLinkedin
peadarcoyle@googlemail.com