Assessments of Resilient Machine Learning Models against Attacks Using Data Poisoning

In the context of artificial intelligence, sustainable Machine learning models play a crucial role for resolving ethical and environmental issues. But their effectiveness depends on how resistant they are to hostile attacks, especially those that contaminate data. In their resilience to data poisoning attacks, the sustainability of machine learning models is examined in this work. We investigate the robustness of different models under different situations of data poisoning through a thorough review. According to our research, resilience features must be included in ML model design and training in order to guarantee the models long-term viability in practical settings