The sensor data obtained from mobile and wearable devices are useful to analyze and estimate user's context, but also user's bio-signals are, because they may reflect user's psychological aspects in the corresponding context. Therefore, in this paper, we focus on context analysis and estimation of mobile users by using bio-signals and sensor data of mobile devices. For the analysis and estimation, various machine learning methods are applied to classify the data into pre-defined six contexts. The evaluation shows that Gradient Boosting Decision Tree achieves the highest classification accuracy of about 80% in supervised methods, and Sparse Representation-based Classification achieves more than 90% accuracy. The results suggest that the context analysis and estimation can be done accurately by using bio-signals and sensor data.