Skip to content

This project applied machine learning to reconstruct activity patterns of employees through the access control information and detected abnormal activities immediately.

Notifications You must be signed in to change notification settings

daniel-code/Access_Control_Patterns_Detection

Repository files navigation

Access Control Patterns Detection

簡介

傳統感應式門禁系統記錄員工刷卡資訊,然而門禁系統僅記錄資訊,必須透過異常規則或人工判讀才能偵測異常門禁行為。 換言之,若有新不當行為模式發生,傳統感應式門禁系統將無法偵測不當行為,必須等到監控人員察覺或該行為被舉發,方能針對該行為模式手動訂定規則進行防範。 因此傳統感應式門禁系統對新型不當行為不僅易發生遺漏情況,且不一定能夠在關鍵時刻偵測不當行為,容易造成門禁系統漏洞。 本專案利用機器學習技術,利用過往刷卡門禁資料建立刷卡行為模型,自動歸納推導異常行為模式,偵測員工每一筆刷卡紀錄,判斷是否為不當刷卡行為,建立全方位且快速門禁行為偵測系統。 該演算法不僅強化傳統門禁系統安全性,並且已在實務中驗證並成功找出異常行為,為企業安全把關。

系統偵測類別

  • 停留時間 (staytime)
  • 刷卡閘門 (path)

About

This project applied machine learning to reconstruct activity patterns of employees through the access control information and detected abnormal activities immediately.

Topics

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published