The need for advanced medicare systems that provide realtime decisions and swift outcomes at earlier stages has emerged as a promising candidate, motivated by the intersection of data streams and potent machine learning techniques. These data streams might join the learning process with evolving mixed features, where it is the requirement for handling the incremental and decremental attribute and instance learning tasks. Here we introduce our proposed solution that copes with the weaknesses in healthcare systems: how to learn real-time patient’s data to make people’s lives advantageous and healthier. Notably, when these new chunks of healthcare data are continuously forthcoming with missing values, redundancies or inconsistencies, data preprocessing is requested. In this paper, a developed incremental and also decremental learning healthcare application is provided based on k-prototypes algorithm and mRMR feature selection technique. It helps to understand new diseases and therapies, and further to advance healthcare monitoring systems and to enhance clinical care based on mRMR feature selection technique. This proposal presents encouraging experimental results compared to the batch k-prototypes method and number of similar incremental healthcare methods. The obtained clusters’ inertia and run time results emphasize the scalability and the performance of our proposed real-time healthcare application.