2nd/10_Wiki/Topics/AI_and_ML/RMSProp-Optimizer.md

id, title, category, status, canonical_id, duplicate_of, aliases, source_trust_level, confidence_score, verification_status, tags, last_reinforced, github_commit

id	title	category	status	canonical_id	duplicate_of	aliases	source_trust_level	confidence_score	verification_status	tags	last_reinforced	github_commit
wiki-2026-0508-rmsprop-optimizer	RMSProp Optimizer	10_Wiki/Topics	duplicate	optimization-in-ai	Optimization-in-AI		A	0.9	redirected	duplicate optimizer rmsprop deep-learning	2026-05-10	pending

RMSProp Optimizer

이 문서는 Optimization-in-AI 의 중복본입니다. Canonical 문서로 redirect.

핵심 요약

• RMSProp (Hinton 2012, Coursera lecture): per-parameter adaptive LR, EMA of squared gradients.
• Update: v ← β·v + (1−β)·g²; θ ← θ − lr·g / (√v + ε).
• 매 Adagrad 의 monotonic decay 문제 의 fix (EMA 의 forget). Adam 의 RMSProp + momentum.
• 매 still used: RNN, RL (DQN classic), 매 modern default 의 AdamW.

🔗 Graph

• 부모: Optimization-in-AI (canonical)

🕓 변경 이력

날짜	변경
2026-05-08	Phase 1
2026-05-10	중복 처리 — canonical 문서로 redirect