Hacker News
new
|
past
|
comments
|
ask
|
show
|
jobs
|
submit
login
zellyn
on Sept 25, 2017
|
parent
|
context
|
favorite
| on:
The Mathematics of 2048: Counting States with Comb...
I thought AlphaGo was still using Monte Carlo tree search, just with two neural networks: one to estimate the board values, and one to estimate likelihood of moves for sampling.
kyleschiller
on Sept 25, 2017
[–]
Sort of, it uses Monto Carlo + a bunch of other things. [0]
[0]
https://medium.com/@karpathy/alphago-in-context-c47718cb95a5
Guidelines
|
FAQ
|
Lists
|
API
|
Security
|
Legal
|
Apply to YC
|
Contact
Search:
