Contents:
0:11 - Resources for this talk
0:30 - Who Am I
1:40 - The Power of Bayesian Statistics
2:21 - Bayesian Statistics for Decision Making
4:48 - An Example: Newspaper Vendor
6:19 - Decision 1: Buys More Newspaper
6:59 - Decision 2: Buys Less Newspaper
8:06 - Profit Curve
9:02 - Scipy Optimizer
11:14 - More Real-World Problem Setting
12:07 - Methodology 1: Use the mean of the observed data
12:36 - Pitfalls of Methodology 1
13:13 - Methodology 2: Let's address uncertainty
15:05 - Summary of Methodology 2
15:28 - Pitfalls of Methodology 2
16:19 - Methodology 3: Let's use Bayesian Statistics
19:29 - Posterior Predictive
19:55 - Recap of Distributions in Bayesian Statistics
20:53 - Optimization
22:15 - Summary of Methodology 3
22:53 - Recap of Three Methodologies
23:22 - Are there other reasons to use Bayesian methods though?
28:30 - Ravin's opinion: Greater community adoption of principled workflow
29:38 - So why are methods like critical fractile still around?
30:47 - When I suggest you use Bayesian methods
31:52 - Recap of we covered
32:19 - Bottom Line: Some decisions are just that important
32:34 - Where to go from here
Ravin Kumar - Bayesian Decision Science https://www.youtube.com/watch?v=WT7o8QRIiQI
Contents: 0:11 - Resources for this talk 0:30 - Who Am I 1:40 - The Power of Bayesian Statistics 2:21 - Bayesian Statistics for Decision Making 4:48 - An Example: Newspaper Vendor 6:19 - Decision 1: Buys More Newspaper 6:59 - Decision 2: Buys Less Newspaper 8:06 - Profit Curve 9:02 - Scipy Optimizer 11:14 - More Real-World Problem Setting 12:07 - Methodology 1: Use the mean of the observed data 12:36 - Pitfalls of Methodology 1 13:13 - Methodology 2: Let's address uncertainty 15:05 - Summary of Methodology 2 15:28 - Pitfalls of Methodology 2 16:19 - Methodology 3: Let's use Bayesian Statistics 19:29 - Posterior Predictive 19:55 - Recap of Distributions in Bayesian Statistics 20:53 - Optimization 22:15 - Summary of Methodology 3 22:53 - Recap of Three Methodologies 23:22 - Are there other reasons to use Bayesian methods though? 28:30 - Ravin's opinion: Greater community adoption of principled workflow 29:38 - So why are methods like critical fractile still around? 30:47 - When I suggest you use Bayesian methods 31:52 - Recap of we covered 32:19 - Bottom Line: Some decisions are just that important 32:34 - Where to go from here