AIセキュリティポータルbot

Towards Deep Learning Models Resistant to Adversarial Attacks

Authors: Aleksander Madry, Aleksandar Makelov, Ludwig Schmidt, Dimitris Tsipras, Adrian Vladu | Published: 2017-06-19 | Updated: 2019-09-04
モデルの頑健性保証
ロバスト性に関する評価
敵対的サンプル

Automated Poisoning Attacks and Defenses in Malware Detection Systems: An Adversarial Machine Learning Approach

Authors: Sen Chen, Minhui Xue, Lingling Fan, Shuang Hao, Lihua Xu, Haojin Zhu, Bo Li | Published: 2017-06-13 | Updated: 2017-10-31
マルウェア分類
毒データの検知
特徴選択手法

Analysis of Anomalies in the Internet Traffic Observed at the Campus Network Gateway

Authors: Veronica del Carmen Estrada | Published: 2017-06-10
TCPセッション管理
データ収集
異常検知アルゴリズム

Certified Defenses for Data Poisoning Attacks

Authors: Jacob Steinhardt, Pang Wei Koh, Percy Liang | Published: 2017-06-09 | Updated: 2017-11-24
ポイズニング
最適化問題
毒データの検知

A Mean-Field Stackelberg Game Approach for Obfuscation Adoption in Empirical Risk Minimization

Authors: Jeffrey Pawlick, Quanyan Zhu | Published: 2017-06-08 | Updated: 2017-10-16
プライバシー保護メカニズム
プライバシー問題
差分プライバシー

Localization of JPEG double compression through multi-domain convolutional neural networks

Authors: Irene Amerini, Tiberio Uricchio, Lamberto Ballan, Roberto Caldelli | Published: 2017-06-06
データフロー解析
モデルアーキテクチャ
画像フォレンジック

Adversarial-Playground: A Visualization Suite for Adversarial Sample Generation

Authors: Andrew Norton, Yanjun Qi | Published: 2017-06-06 | Updated: 2017-06-16
モデルの頑健性保証
攻撃タイプ
敵対的学習

Context-aware, Adaptive and Scalable Android Malware Detection through Online Learning (extended version)

Authors: Annamalai Narayanan, Mahinthan Chandramohan, Lihui Chen, Yang Liu | Published: 2017-06-03 | Updated: 2017-07-06
オンライン学習
プライバシー問題
特徴選択

MagNet: a Two-Pronged Defense against Adversarial Examples

Authors: Dongyu Meng, Hao Chen | Published: 2017-05-25 | Updated: 2017-09-11
攻撃タイプ
敵対的サンプルの検知
防御手法の効果分析

Detecting Malignant TLS Servers Using Machine Learning Techniques

Authors: Sankalp Bagaria, R. Balaji, B. S. Bindhumadhava | Published: 2017-05-25
プロトコル手順
悪意のあるクライアント
特徴選択