dremmettbrown

5 ms·

1.
by dremmettbrown 4y ago
Parallel Markets | Sr Frontend Engineer | Remote | Full Time | https://parallelmarkets.com We're looking for a very special someone to lead our frontend team over at Parallel Markets, a fintech startup with solid traction a
2.
by dremmettbrown 4y ago
Parallel Markets | Sr Frontend Engineer | Remote | Full Time | https://parallelmarkets.com Parallel Markets provides a re-usable identity for financial institutions in both web2 and web3 contexts. We just raised our Series A fro
3.
by dremmettbrown 5y ago
Parallel Markets | Remote | Full Time | https://parallelmarkets.com Parallel Markets provides a re-usable identity for financial institutions in both web2 and web3 contexts. We just raised our Series A from Union Square Ventures
4.
The role of fear in a startup(findingscience.com)
2 points by dremmettbrown 14y ago|0 comments
5.
by dremmettbrown 14y ago
That ignores the effect of your friends sharing with their friends and so on. I added the final graph to illustrate this point - yes, having 2 entries is better than having 1, but the network effect of your friends pulling in their friends
6.
by dremmettbrown 14y ago
zck is right about failing to account for the friends votes - which does mean an asymptotic approach to 0.5 rather than 1. I updated the graph (and the example fractions). That does, however, change the conclusion, that you shouldn't shar
7.
The Paradox of Sharing: should you share a contest if you get extra entries?(findingscience.com)
17 points by dremmettbrown 14y ago|12 comments
8.
by dremmettbrown 14y ago
"A/B testing is meant for strict experiments where focus is on statistical significance, whereas multi-armed bandit algorithms are meant for continuous optimization where focus is on maintaining higher average conversion rate." Who cares ab
9.
by dremmettbrown 14y ago
https://github.com/bmuller/bandit