Machine Learning techniques on different sectors plays a vital role for prediction analysis. In healthcare sector, diagnosis support and prognostic evaluation has been done using various algorithms of machine learning. The Multiple Sclerosis (MS) is one of the chronic diseases which affects the central nervous system of human. There are multiple symptoms and disabilities indicate the issues of Multiple Sclerosis. This neurological disorder has various patterns and disease progression rate among worldwide. This causes a variety of symptoms including fatigue, numbness, visual difficulties, urinary incontinence, and difficulty moving and cognitive deficiencies. This paper analyzes the risk factors of Multiple Sclerosis and early prediction using Machine learning methods. This work proposed hybrid Recurrent Neural Network with mini-batch gradient descent method for Multiple Sclerosis prediction. The performance of this proposed method is evaluated and the achieved accuracy rate is 92%.