scikit-learn/scikit-learn

Consensus Index for Clustering Model Selection

Open

#11,778 opened on 2018年8月7日

GitHub で見る
 (6 comments) (1 reaction) (0 assignees)Python (66,084 stars) (27,020 forks)batch import
Documentationhelp wantedmodule:cluster

説明

The sentence:

XXX can also be useful in a purely unsupervised setting as a
  building block for a Consensus Index that can be used for clustering
  model selection (TODO).

is in the description of every clustering metric that requires a ground truth label. Is there a write-up somewhere of an example workflow to build such a Consensus Index using these metrics? I searched through old issues and the original PRs creating this documentation and I can't see anything that looks like that. Though to be fair, I'm a big dummy, so I could just be missing it. If there isn't, is this TODO still really a TODO and is there an existing issue for it that I'm also too dumb to find? Thanks!

❤️

cc @ogrisel since you seem to be the og implementor and the original author of these docs.

コントリビューターガイド

技術スタック
python
領域
documentationmachine learning
Issue 種別
documentation
難度新しい貢献者にとっての実装難度です。1 は非常に小さな変更、5 は専門的な作業です。
2
推定時間経験ある貢献者が調査、実装、テスト、pull request 準備にかけるおおよその時間範囲です。
1-3 hours
活動状況Issue が今どれだけ取り組みやすいかを示します。新しい、活発、古い、ブロック中、またはメンテナー入力待ちなどです。
stale
明確さ期待される変更、受け入れ条件、次の手順がどれだけ明確かを示します。
clear
前提条件
basic understanding of clustering metrics
初心者向け度初回貢献者にどれだけ取り組みやすいかを 1-100 で推定したスコアです。
70
調査方針
Look at the scikit learn documentation for clustering metrics that mention 'Consensus Index'. Search for any existing write up or example workflow in the repository. Check the linked PRs or original author comments. Consider creating a new example notebook or adding a note in the documentation. If no prior work exists, outline a simple workflow using metrics like adjusted Rand index for model selection.