To utilize the Machine Learning tools in early diagnosis of sepsis for better treatment
A Sepsis is a serious condition resulting from the presence of harmful synthetics in blood or other tissues and the body’s response to their presence, potentially leading to the malfunctioning of various organs, shock and death. Sepsis is activated by the immune system present in our body that works all the time in order to prevent infection from entering into body. During this stage, the enormous number of synthetic substances discharged into the blood causes broad irritation. For the patient the practicality of detecting sepsis disease occurrence in development is an important factor in the result. The primary goal of our project is to build, train and test a Multi-Layer Perceptron (MLP) model. Using data that is available in the form of electronic clinical health data. We can train the system to predict outcome of class labels as sepsis or no-sepsis for new health records. The early prediction of sepsis is potentially life-saving, and we aim to predict sepsis 6 hours before the clinical prediction of sepsis.
- Collected the clinical datafiles and it has been converted from pipe separated file to comma separated file and with the help of resampling the data have been balanced and saved on CSV file for easy processing.
- Research about Machine Learning Algorithms suitable for dataset.
- Trained model with different algorithms and taken the best accuracy model.
- Generated pickel file for production purpose.
- Developed User Interface.
- Deployed on herokuapp
- Developing RESTAPI.
- Production on GoDaddy.