• Shreeya P
• Vanishri Sataraddi

Larger datasets are required when machine learning systems grow, which forces practitioners to depend on computerized and externalized data gathering techniques to be able to stay competitive. There are serious security dangers associated with moving out of direct, reliable human control, including the possibility of training data modification intended to change or compromise the behavior of the emerging models. This survey intends to outline the spectrum of prospective exploits, thoroughly classify and investigate multiple weaknesses inside data set collection generation and utilization, and evaluate defensive tactics contrary to such risks. It also draws attention to open problems in the industry and urges greater study and advancement to strengthen predictive systems contrary to such changing security threats.

"Investigating Machine Learning's Dataset Security Issues: A Comprehensive Analysis of Data Intoxication, Backdoor assaults, and Potential Defense Strategies", International Journal of Emerging Technologies and Innovative Research (www.jetir.org), ISSN:2349-5162, Vol.11, Issue 6, page no.c50-c55, June-2024, Available :http://www.jetir.org/papers/JETIR2406206.pdf

"Investigating Machine Learning's Dataset Security Issues: A Comprehensive Analysis of Data Intoxication, Backdoor assaults, and Potential Defense Strategies", International Journal of Emerging Technologies and Innovative Research (www.jetir.org | UGC and issn Approved), ISSN:2349-5162, Vol.11, Issue 6, page no. ppc50-c55, June-2024, Available at : http://www.jetir.org/papers/JETIR2406206.pdf