20120221 asa chicago_chapter
Post on 26-Jan-2015
109 Views
Preview:
DESCRIPTION
Transcript
@deanmalmgren
2012 februarychicago chapter asa
@deanmalmgren
2012 februarychicago chapter asa
data-driventhe love child of statistics & design
chicago chapter asa | @deanmalmgren | 2012 february
analytics
consulting design
chicago chapter asa | @deanmalmgren | 2012 february
data-driven innovation
data appraisaldiscovering value from shipwrecked data
what data do you have?
how can it be used?
market research and internal testingfinding the Jeremy Lin on your bench
what is feasible?
what adds value?
prototype developmentfrom blueprints to concept car
how will it work?
how will it be used?
product implementation scalable solutions, understandable by humans
how will it be maintained?
what are technology limitations?
chicago chapter asa | @deanmalmgren | 2012 february
data appraisaldiscovering value from shipwrecked data
market research and internal testingfinding the Jeremy Lin on your bench
prototype developmentfrom blueprints to concept car
product implementation scalable solutions, understandable by humans
implementation of prototype vastly expanded market size
data-driven interface designglobal information services provider
rapidly developed customer-facing prototype interface for market testing
chicago chapter asa | @deanmalmgren | 2012 february
data appraisaldiscovering value from shipwrecked data
market research and internal testingfinding the Jeremy Lin on your bench
prototype developmentfrom blueprints to concept car
product implementation scalable solutions, understandable by humans
designed analytics engine to semi-automatically classify documents for legal relevance
analytics engine decreased costs by >70%, dramatically changing their business model
data-driven analytics engineglobal e-discovery solutions provider
chicago chapter asa | @deanmalmgren | 2012 february
data appraisaldiscovering value from shipwrecked data
market research and internal testingfinding the Jeremy Lin on your bench
prototype developmentfrom blueprints to concept car
product implementation scalable solutions, understandable by humans
data-driven change managementprocter & gamble
developed methodology for identifying optimal cohort of change ambassadors
models predict increased speed of adoption by 50% compared to managerʼs picks
chicago chapter asa | @deanmalmgren | 2012 february
change management in a nutshell
chicago chapter asa | @deanmalmgren | 2012 february
change management in a nutshell
chicago chapter asa | @deanmalmgren | 2012 february
change management in a nutshell
chicago chapter asa | @deanmalmgren | 2012 february
change management in a nutshell
not enough change
chicago chapter asa | @deanmalmgren | 2012 february
change management in a nutshell
not enough change“ ”do not trust “ ”
chicago chapter asa | @deanmalmgren | 2012 february
build trust with “organic” ambassador picks
chicago chapter asa | @deanmalmgren | 2012 february
social network survey
build trust with “organic” ambassador picks
chicago chapter asa | @deanmalmgren | 2012 february
social network survey
build trust with “organic” ambassador picks
contagion modeling
chicago chapter asa | @deanmalmgren | 2012 february
social network survey
build trust with “organic” ambassador picks
contagion modelingambassador team optimization
chicago chapter asa | @deanmalmgren | 2012 february
social network surveyidentify mavens and fill in the gaps
chicago chapter asa | @deanmalmgren | 2012 february
social network surveyidentify mavens and fill in the gaps
chicago chapter asa | @deanmalmgren | 2012 february
social network surveyidentify mavens and fill in the gaps
chicago chapter asa | @deanmalmgren | 2012 february
social network surveyidentify mavens and fill in the gaps
ask respondents who they trust
chicago chapter asa | @deanmalmgren | 2012 february
social network surveyidentify mavens and fill in the gaps
ask respondents who they trust
chicago chapter asa | @deanmalmgren | 2012 february
social network surveyidentify mavens and fill in the gaps
ask respondents who they trust
chicago chapter asa | @deanmalmgren | 2012 february
social network surveyidentify mavens and fill in the gaps
ask respondents who they trust
chicago chapter asa | @deanmalmgren | 2012 february
social network surveyidentify mavens and fill in the gaps
? ? ? ? ? ? ? ? ? ? ? ? ?? ? ? ? ? ? ? ? ? ? ? ? ?
? ? ? ? ? ? ? ? ? ? ? ? ?? ? ? ? ? ? ? ? ? ? ? ? ?
? ? ? ? ? ? ? ? ? ? ? ? ?? ? ? ? ? ? ? ? ? ? ? ? ?
? ? ? ? ? ? ? ? ? ? ? ? ?? ? ? ? ? ? ? ? ? ? ? ? ?
ask respondents who they trust
chicago chapter asa | @deanmalmgren | 2012 february
social network surveyidentify mavens and fill in the gaps
ask respondents who they trust
use node attributes to estimate connectivity
chicago chapter asa | @deanmalmgren | 2012 february
social network surveyidentify mavens and fill in the gaps
ask respondents who they trust
use node attributes to estimate connectivity
chicago chapter asa | @deanmalmgren | 2012 february
susceptible
infected
contagion modelingsi models, modified by iphone, averaged like ipcc
chicago chapter asa | @deanmalmgren | 2012 february
susceptible
infected
contagion modelingsi models, modified by iphone, averaged like ipcc
prob
abili
ty in
fect
ed0 1
0
1
fraction of infected neighbors
chicago chapter asa | @deanmalmgren | 2012 february
susceptible
infected
contagion modelingsi models, modified by iphone, averaged like ipcc
chicago chapter asa | @deanmalmgren | 2012 february
susceptible
infected
contagion modelingsi models, modified by iphone, averaged like ipcc
chicago chapter asa | @deanmalmgren | 2012 february
susceptible
infected
contagion modelingsi models, modified by iphone, averaged like ipcc
fraction of infected neighbors
prob
abili
ty in
fect
ed0 1
0
1
chicago chapter asa | @deanmalmgren | 2012 february
susceptible
infected
contagion modelingsi models, modified by iphone, averaged like ipcc
fraction of infected neighbors
prob
abili
ty in
fect
ed0 1
0
1
chicago chapter asa | @deanmalmgren | 2012 february
susceptible
infected
contagion modelingsi models, modified by iphone, averaged like ipcc
fraction of infected neighbors
prob
abili
ty in
fect
ed0 1
0
1
uncertain about curve shape?use lots of models and simulations!
chicago chapter asa | @deanmalmgren | 2012 february
ambassador team optimizationnavigate rough energy landscape
chicago chapter asa | @deanmalmgren | 2012 february
ambassador team optimizationnavigate rough energy landscape
time
num
ber i
nfec
ted
0 !0
n
chicago chapter asa | @deanmalmgren | 2012 february
ambassador team optimizationnavigate rough energy landscape
time
num
ber i
nfec
ted
0 !0
n
chicago chapter asa | @deanmalmgren | 2012 february
ambassador team optimizationnavigate rough energy landscape
time
num
ber i
nfec
ted
0 !0
n
chicago chapter asa | @deanmalmgren | 2012 february
ambassador team optimizationnavigate rough energy landscape
time
num
ber i
nfec
ted
0 !0
n
chicago chapter asa | @deanmalmgren | 2012 february
ambassador team optimizationnavigate rough energy landscape
time
num
ber i
nfec
ted
0 !0
n
chicago chapter asa | @deanmalmgren | 2012 february
ambassador team optimizationnavigate rough energy landscape
time
num
ber i
nfec
ted
0 !0
n
250064( ( simulated
annealing=
chicago chapter asa | @deanmalmgren | 2012 february
data appraisaldiscovering value from shipwrecked data
market research and internal testingfinding the Jeremy Lin on your bench
prototype developmentfrom blueprints to concept car
product implementation scalable solutions, understandable by humans
data-driven change managementprocter & gamble
http://acme.datascopeanalytics.com
top related