We do this in RL, it's why there's a temporal discount factor for reward.



The further out you extrapolate the more you have to discount your prediction due to compounding uncertainty.

Our brains are hard wired like this for a fundamental reason.
WHY5.82%
MORE0.87%
This page may contain third-party content, which is provided for information purposes only (not representations/warranties) and should not be considered as an endorsement of its views by Gate, nor as financial or professional advice. See Disclaimer for details.
  • Reward
  • 7
  • Share
Comment
0/400
OnChainDetectivevip
· 08-04 20:49
hmm interesting pattern... statistically speaking the neural wiring matches typical reward decay curves
Reply0
GasFeeThundervip
· 08-04 15:53
Data shows that yield decay is inevitable.
View OriginalReply0
SingleForYearsvip
· 08-03 02:04
Forecasting also needs to consider the pros and cons of short and long term.
View OriginalReply0
Blockwatcher9000vip
· 08-03 01:58
Ah? This theory is actually quite interesting.
View OriginalReply0
NotAFinancialAdvicevip
· 08-03 01:56
Indeed, there aren't that many future matters.
View OriginalReply0
StopLossMastervip
· 08-03 01:43
Don't do things that are too difficult; just do something simple and it's very stable.
View OriginalReply0
VibesOverChartsvip
· 08-03 01:42
so based fr
Reply0
Trade Crypto Anywhere Anytime
qrCode
Scan to download Gate app
Community
English
  • 简体中文
  • English
  • Tiếng Việt
  • 繁體中文
  • Español
  • Русский
  • Français (Afrique)
  • Português (Portugal)
  • Bahasa Indonesia
  • 日本語
  • بالعربية
  • Українська
  • Português (Brasil)