Slide 26
Slide 26 text
GitHub: codess-aus
Twitter: codess_aus
Dogs: Leo the Lion, Snickers
Role: Developer Engagement PM for Data
Science and Data Engineering
Degrees: MSC Comp Sci + BA
Philosophy Hons
Employer: Microsoft
Languages: English, French
Pronouns: She/Her
Age: GenX
Weight: Over
Mentors: She Codes, Muses
Code JS, GPN, ABCN
Codes: Python, JavaScript
Speaker: 0.67
Boards: 0.54
Tag: 0.98
Speaker: Tedx Perth, DDDPerth
Boards: ACS Chair,
FutureNow Director
Tag: Michelle @ Microsoft
@jkeatnuxsuo #DDDAzure @Codess_aus
J: What about you, in your case, those things are not the most important features
M: Exactly, I always said community advocate even when I was working as a Service
Delivery Manager – because although SDM was the job title I had, it didn’t describe
who I was.
J: There is data missing from your dataset too.
M: Yes, but it’s not significant for anything I want to do with my life, and if those
features took on a higher weight, it would definitely reflect bias in the dataset.
J: So, when different features have a different priority in a particular instance from what
is globally perceived to be important, we call that Local Feature importance.
M: you got it
25