From f106df191e036dd4947edbd4b1b9de56b1f3276c Mon Sep 17 00:00:00 2001 From: smilesun Date: Thu, 17 Oct 2024 07:03:06 +0200 Subject: [PATCH] update doc --- docs/docIRM.md | 3 ++- docs/docMA.md | 3 ++- docs/doc_coral.md | 2 +- docs/doc_irl.md | 2 +- docs/index.rst | 1 + 5 files changed, 7 insertions(+), 4 deletions(-) diff --git a/docs/docIRM.md b/docs/docIRM.md index fd097e2aa..43c93d355 100644 --- a/docs/docIRM.md +++ b/docs/docIRM.md @@ -1,4 +1,5 @@ -# Trainer IRM (Invariant Risk Minimization) +# Trainer IRM +## Invariant Risk Minimization Decompose a classification task into feature extraction $\Phi(\cdot)$ and classificaiton layer $w(\cdot)$, then the task loss is diff --git a/docs/docMA.md b/docs/docMA.md index 991f5ad77..1f6c4b51a 100644 --- a/docs/docMA.md +++ b/docs/docMA.md @@ -1,4 +1,5 @@ -# Simple Moving Average +# Trainer SMV +## Simple Moving Average For each epoch, convex combine the weights for each layey from Paper: Ensemble of Averages: Improving Model Selection and Boosting Performance in Domain Generalization, Devansh Arpit, Huan Wang, Yingbo Zhou, Caiming Xiong, Salesforce Research, USA diff --git a/docs/doc_coral.md b/docs/doc_coral.md index a79be02b7..6df423f2e 100644 --- a/docs/doc_coral.md +++ b/docs/doc_coral.md @@ -1,4 +1,4 @@ -# Deep CORAL +# Trainer CORAL ## Deep CORAL: Correlation Alignment for Deep Domain Adaptation nonlinear transformation that aligns correlations of diff --git a/docs/doc_irl.md b/docs/doc_irl.md index 46a907314..7c5204ab2 100644 --- a/docs/doc_irl.md +++ b/docs/doc_irl.md @@ -1,4 +1,4 @@ -# Causal IRL +# Trainer Causal IRL ## Invariant Causal Mechanisms through Distribution Matching Source: https://arxiv.org/pdf/2206.11646 diff --git a/docs/index.rst b/docs/index.rst index d6226f5c9..2d2559de1 100644 --- a/docs/index.rst +++ b/docs/index.rst @@ -24,6 +24,7 @@ Welcome to domainlab's documentation! docMA doc_coral doc_irl + doc_miro Indices and tables