Hyperparameters are important in machine learning because they affect how well models perform. Some hyperparameters are kept secret because they are valuable for businesses. This work shows that it is possible to steal these hyperparameters from machine learning models. The attacks work on popular algorithms like ridge regression, logistic regression, support vector machine, and neural networks. The researchers tested their attacks on Amazon Machine Learning and found that they could accurately steal hyperparameters. This study suggests that new defenses are needed to protect against these attacks for certain machine learning algorithms.