yzhao062/pyod

Support for DBSCAN based methods

Open

#303 建立於 2021年5月12日

在 GitHub 查看
 (1 留言) (0 反應) (0 負責人)Python (7,762 star) (1,308 fork)batch import
help wanted

描述

Is it possible to add DBSCAN based methods (https://en.wikipedia.org/wiki/DBSCAN#Extensions) for anomaly detection?

The normal DBSCAN/OPTICS in sklearn don't support a predict method as such but the hdbscan library supports one.

貢獻者指南

技術棧
pythonscikit learn
領域
machine learningdata
議題類型
feature
難度面向新貢獻者的預計實作難度，1 表示很小改動，5 表示專家級工作。
3
預計時間有經驗貢獻者完成調查、實作、測試並準備 pull request 的粗略時間範圍。
3-5 days
活動狀態議題目前的可參與程度：新鮮、活躍、陳舊、阻塞或等待維護者輸入。
stale
清晰度議題是否清楚說明預期改動、驗收標準和下一步。
mostly clear
前置要求
Basic knowledge of DBSCANPythonscikit learn
新手友善度1-100 的估計分數，表示該議題對首次貢獻者的友善程度。
30
研究方向
To implement DBSCAN based methods, study the hdbscan library's predict functionality compared to sklearn's DBSCAN. Integrate the algorithm into pyod by adding a new detector class following existing conventions, inheriting from BaseDetector, and implementing fit and predict methods. Review pyod/models for structure and test with standard outlier detection datasets.