dataRLsec: Safety, Security, and Reliability With Robust Offline Reinforcement Learning for DPAs

TL;DR

This paper investigates data poisoning attacks on AI systems, analyzes existing defenses, and proposes a robust offline reinforcement learning approach with weighted hash verification and density-ratio weighted behavioral cloning to enhance safety and security.

Contribution

It introduces a novel offline RL framework with four-stage algorithms and combines DWBC with data defense strategies to counter poisoning attacks in AI training data.

Findings

Analyzed risks of DPAs in AI systems.

Proposed a four-stage offline RL algorithm for robustness.

Demonstrated improved safety and security in experiments.

Abstract

Data poisoning attacks (DPAs) are becoming popular as artificial intelligence (AI) algorithms, machine learning (ML) algorithms, and deep learning (DL) algorithms in this artificial intelligence (AI) era. Hackers and penetration testers are excessively injecting malicious contents in the training data (and in testing data too) that leads to false results that are very hard to inspect and predict. We have analyzed several recent technologies used (from deep reinforcement learning to federated learning) for the DPAs and their safety, security, & countermeasures. The problem setup along with the problem estimation is shown in the MuJoCo environment with performance of HalfCheetah before the dataset is poisoned and after the dataset is poisoned. We have analyzed several risks associated with the DPAs and falsification in medical data from popular poisoning data attacks to some popular data defenses. We have proposed robust offline reinforcement learning (Offline RL) for the safety and reliability with weighted hash verification along with density-ratio weighted behavioral cloning (DWBC) algorithm. The four stages of the proposed algorithm (as the Stage 0, the Stage 1, the Stage 2, and the Stage 3) are described with respect to offline RL, safety, and security for DPAs. The conclusion and future scope are provided with the intent to combine DWBC with other data defense strategies to counter and protect future contamination cyberattacks.