Updates
(latest listed first)
- Lectures: Lace2; DE Stats; 6 learners; No free lunch; Decision trees; One;
- Review: 4; 3; 2; 1
- War stories: And the best report is…; When models kill;
- Tutorials: Weka101;
- Homeworks: Submit here; Aug-Sept
- Project: Essay
NCSU CSC 591-023 (10722)
CSC 791-023 (11162). Tues/Thurs 4:30 to 5:45
Mid-session exam week 10 (on terminology) No final exam.
An important part of knowledge is “no”; i.e. the ability to critically assess something, and to recognize when one idea is better, smarter, than another.
“Science” is the process of communities sharing and reviewing and improving each other’s ideas. Sadly, most “data science” is not about “science”. Rather its about vendors selling you stuff that does not work properly and does not ring an alarm when it starts failing.
So a million million people can run data miners. But how many now when those data miners start going wrong? And how to fix faulty models?
Do this subject, learn answers to those questions, become the data scientist everyone needs to consult with when things start going wrong.