The Chest X-Ray imaging is one of the most common medical imaging field which even today relies mostly on the expert knowledge and careful manual examination. But classification of X-Ray disease into one of thoracic classes is one of the most challenging task because these diseases happen in localized disease specific area and sometimes even for the expert radiologists it is very difficult to identify the disease in short span of time. Hence there is a need to introduce some efficient models which can extract the latent features to ease this task of classification.With the availability of large sized dataset of Chest X-Ray images which have been released by the NIH Health Institute, it is now possible for researchers across the globe to create a model which can classify the disease present in chest X-Ray images into thoracic classes and can help the radiologist in identifying the disease in short span of time.Through this research we propose a supervised learning model a model which can perform multi label chest X-Ray image classification with reduced dimensionality of X-Ray images to overcome the above mentioned limitations.