tysonzni

6 ms·

1.
Statistical Paradises and Paradoxes in Big Data [pdf](statistics.fas.harvard.edu)
3 points by tysonzni 7y ago|0 comments
2.
Contagion in Mass Killings and School Shootings (2015)(journals.plos.org)
72 points by tysonzni 7y ago|118 comments
3.
Curiosity Detects Unusually High Methane Levels(jpl.nasa.gov)
2 points by tysonzni 7y ago|0 comments
4.
Why Data Science Teams Need Generalists, Not Specialists(hbr.org)
5 points by tysonzni 8y ago|0 comments
5.
by tysonzni 8y ago
I think it makes sense to ask companies to explicitly list cultural values they prioritize the most as well as what they value the least . Evaluating/creating a company culture is in a way really about understanding what the tradeoffs
6.
by tysonzni 8y ago
Related wikipedia article: https://en.wikipedia.org/wiki/The_Elephant%27s_Foot
7.
High-Speed Rail’s Central Valley Segment Is Not Stranded(sf.streetsblog.org)
1 points by tysonzni 8y ago|3 comments
8.
Student Loan Relief or Paid Vacation? These Workers Get a Choice(bloomberg.com)
2 points by tysonzni 8y ago|0 comments
9.
Business secrets from terrible people(erikbern.com)
3 points by tysonzni 8y ago|0 comments
10.
The Nuclear Power Plant of the Future May Be Floating Near Russia(nytimes.com)
96 points by tysonzni 8y ago|115 comments
11.
Augur Prediction Market Data(predictions.global)
2 points by tysonzni 8y ago|0 comments
12.
Walmart's Fintech Partner Helps Break Paycheck-To-Paycheck Cycle(bloomberg.com)
7 points by tysonzni 8y ago|1 comments
13.
by tysonzni 8y ago
Is the game for members only?
14.
by tysonzni 8y ago
No it does not. From NIH study: ( https://www.ncbi.nlm.nih.gov/pmc/articles/PMC5026916/ ) >We use high-quality linked Social Security and Medicare records to estimate life tables for the older U.S. populatio
15.
by tysonzni 8y ago
I don't know how China collect its public health statistics, but its economics numbers have been extensively studied. And the consensus is that its official GDP growth figures were overstated historically (by as much as 67%), but had b
16.
TensorFlow for R(tensorflow.rstudio.com)
10 points by tysonzni 8y ago|0 comments
17.
Smart Compose: Using Neural Networks to Help Write Emails(ai.googleblog.com)
3 points by tysonzni 8y ago|0 comments
18.
by tysonzni 8y ago
Yeah you're right. Randomization at least ensures internal validity of the study as applied to people who are interested in WFH . Whether it generalizes to others is an open question.
19.
As flat prices soar in Hong Kong, investors park money down in the lot(reuters.com)
1 points by tysonzni 8y ago|0 comments
20.
by tysonzni 8y ago
The paper mentioned how randomization (a lottery draw on odd/even birth dates) was done among employees who expressed interest. So it is a randomized experiment.
21.
by tysonzni 8y ago
According to the paper, the random selection was done on participants who expressed interest. > Approximately half of the employees (503) were interested, particularly those who had less education and tenure, their own rooms, and faced
22.
by tysonzni 8y ago
In the new equilibrium of people sending lots of automated emails, how would the receiver of an email distinguish between a canned response and a real response? I suspect the incentive to properly signal your email as real can reduce adopti
23.
by tysonzni 8y ago
You're right. But I think the single-value, hyper-transparent nature of prediction market can create an institutional feedback loop that's more powerful than what arises from the diverse, individually held opinions that others don
24.
by tysonzni 8y ago
How do you deal with the cultural challenge of convincing people there's value and selfish benefits to having more transparency? I've noticed that in large organizations certain information are held by a small number of people who
25.
Interviewing is a noisy prediction problem(erikbern.com)
2 points by tysonzni 8y ago|1 comments
26.
Unlocking the Business Secrets of Escape Rooms(wsj.com)
3 points by tysonzni 8y ago|0 comments
27.
by tysonzni 8y ago
Interesting point - I highly doubt the suspect was a "prior suspect". But in this case I believe investigators were able to find additional evidence after they matched the DNA. So in Bayesian terms, multiple updates (DNA match + l
28.
by tysonzni 8y ago
It sounds like Peter Thiel's argument about how natural monopolies can afford to invest heavily in R&D because they're financially comfortable. And how competitive industries supposedly don't offer the same room for resea
29.
by tysonzni 8y ago
Increasing returns to scale isn't the only way unit costs can go down. Competitive forces typically incentivize companies to come up with new ways to increase productivity and reduce cost. Having even one competitor is enough to meanin
30.
by tysonzni 8y ago
Lots of people in the DNA databases are relatives to someone who lived in Sacramento. The lead came from the police searching through DNA databases which identified some set of relatives who were suspicious, but they later obtained the DN

More ›