This paper presents a comprehensive survey on detecting human stress levels using sleeping habits and machine learning algorithms. Stress is a significant factor that affects overall health, and poor sleep quality is often linked to elevated stress levels. This survey reviews existing techniques, focusing on the application of machine learning models like Support Vector Machines (SVM) and Random Forest for stress detection. By analyzing sleep data, including sleep duration, snoring range, and REM cycles, these models predict stress levels with considerable accuracy. The results from the review indicate that machine learning-based systems can provide effective early stress detection, leading to timely interventions for better mental and physical health. Additionally, this paper discusses the datasets, features, and evaluation metrics used in various studies, highlighting the strengths and limitations of different approaches.