'

Data Science in Digital Health

Понравилась презентация – покажи это...





Слайд 0

DATA SCIENCE In digital health #wwdh @neal_Lathia


Слайд 1

SOCIAL SCIENCE BEHAVIOUR CHANGE? HUMAN COMPUTER INTERACTION DATA SCIENCE


Слайд 2

AUTOMATE UNDERSTAND BEHAVIOUR CHANGE? DESIGN


Слайд 3

HOW DOES BEHAVIOUR CHANGE? BEHAVIOUR CHANGE? HOW Do PEOPLE INTERACT WITH TECHNLOGY? HOW COULD TECHNLOGY INTERACT WITH PEOPLE?


Слайд 4

DATA SCIENCE DIGITAL BEHAVIOUR CHANGE HUMAN COMPUTER INTERACTION


Слайд 5

CASE 1: memory & choice (NOT HEALTH) MAKING CHOICES


Слайд 6

“Psychologists have recognized for many years that humans have a limited capacity to store current information in memory.” - “Information Overload” on Wikipedia


Слайд 7

SURROUNDED BY CHOICES


Слайд 8

AUTOMATED BY RECOMMENDATION - Neal's slides during his PhD


Слайд 9

AUTOMATED BY RECOMMENDATION Navigating choice ~ Predicting missing data Ranking on predictions - Neal's slides during his PhD


Слайд 10

AUTOMATED BY RECOMMENDATION No “framework” No “item” context No theory/categorisation Simplistic assumption No uniformity 1000 outcomes for 1000 people - Neal's slides during his PhD


Слайд 11

Online Recommendations USES BEHAVIOURAL THEORY EXPLAINS THE BEHAVIOUR AUTOMATED PROCESS ENHANCES ENGAGEMENT CHANGES BEHAVIOUR ALWAYS GETS IT RIGHT NO NO / BADLY YES YES YES NO


Слайд 12

Online Recommendations USES BEHAVIOURAL THEORY EXPLAINS THE BEHAVIOUR AUTOMATED PROCESS ENHANCES ENGAGEMENT CHANGES BEHAVIOUR ALWAYS GETS IT RIGHT NO NO YES YES YES NO DOMAIN KNOWLEDGE DATA SCIENCE BOTH


Слайд 13

“Your decades of specialist knowledge are not only useless, they're actually unhelpful; your sophisticated techniques are worse than generic methods; The algorithms tell you what's important and what's not.. ” - @jeremyphoward (Interview)


Слайд 14

“.. You might ask why those things are important, but I think that's less interesting. You end up with a predictive model that works.” - @jeremyphoward (Interview)


Слайд 15

SOCIAL SCIENCE.. ?


Слайд 16

The Emotion Sense Platform: WHAT SMARTPHONES CAN SENSE THEMSELVES Location, mobility, sociability, physical activity What SMARTPHONES CAN PROMPT YOU TO TELL Mood, symptoms, assessments


Слайд 17


Слайд 18

CASE 2: Automating support QUITTING SMOKING


Слайд 19

Smoking Cessation – Ideal YOUR SMOKING BEHAVIOUR + “ReCOMMENDED” SUPPORT = BEHAVIOUR CHANGE


Слайд 20

Smoking Cessation – Ideal YOUR SMOKING BEHAVIOUR NO DATA ON THE “USER” + “RECOMMENDED” SUPPORT WHAT IS THE “ITEM?” = BEHAVIOUR CHANGE NOT POSSIBLE?


Слайд 21

“Cold start is a potential problem in computer-based information systems (.. WHERE..) the system cannot draw any inferences for users (or items) about which it has not yet gathered sufficient information.” - “Cold Start” on Wikipedia


Слайд 22

And beyond: in a given health domain, what information should we (can we) collect? “Cold start is a potential problem in computer-based information systems (.. WHERE..) the system cannot draw any inferences for users (or items) about which it has not yet gathered sufficient information.” - “Cold Start” on Wikipedia


Слайд 23

HEALTH /SOCIAL SCIENCE Cold start DATA SCIENCE DIGITAL BEHAVIOUR CHANGE HUMAN COMPUTER INTERACTION


Слайд 24

“cue-induced cravings: intense, episodic cravings typically provoked by situational cues associated with drug use (.. ) smokers exposed to smoking-related cues demonstrate increased craving (.. ).” - Ferguson, Shiffman. The relevance and treatment of cue-induced cravings in tobacco dependence. In J Subst Abuse Treat. April 2009.


Слайд 25

“cue-induced cravings: intense, episodic cravings typically provoked by situational cues associated with drug use (.. ) smokers exposed to smoking-related cues demonstrate increased craving (.. ).” Situation: mood, craving, location, social setting - Ferguson, Shiffman. The relevance and treatment of cue-induced cravings in tobacco dependence. In J Subst Abuse Treat. April 2009.


Слайд 26


Слайд 27

EXAMPLE Your location + your profile = tailored support


Слайд 28

Smoking Cessation USES BEHAVIOURAL THEORY EXPLAINS THE BEHAVIOUR AUTOMATED PROCESS ENHANCES ENGAGEMENT CHANGES BEHAVIOUR ALWAYS GETS IT RIGHT YES NO (BUT what DATA!) YES YES? YES? NO


Слайд 29


Слайд 30

AND FINALLY: GOING FORWARD


Слайд 31

UNDERSTAND IMPLEMENT Design Automate EVALUATE HYPOTHESIS Linear/hypothesis driven research: good for publication, bad for software.


Слайд 32

SOFTWARE IS NEVER FINISHED.. . . IT IS UPDATED. MONITOR DELIVER N. Lathia et. al. In IEEE Pervasive Computing. 2013. LEARN HYPOTHESIS


Слайд 33

AUTOMATE UNDERSTAND BEHAVIOUR CHANGE? DESIGN


Слайд 34

SCHIZOPHRENIA FREEMIUM ANXIETY MOOD ADJUSTMENT ANTI-SOCIAL PERSONALITY ON/oFFLINE MOOD EXPRESSION Code: http://emotionsense.github.io/


Слайд 35

DATA SCIENCE In digital health #wwdh @neal_Lathia


Слайд 36


×

HTML:





Ссылка: