The machine learning model analyses the Clinical, Anthropometric and Biochemical(CAB) data to predict the illness-type in children under the age of five years.
To supplement the information provided by Annual Health Survey (AHS), a biomarker component has been introduced in AHS to collect data on nutritional status, life style diseases like diabetes & hypertension and anaemia in Empowered Action Group (EAG) States & Assam. This component, namely Clinical, Anthropometric and Bio-chemical (CAB) survey, is conducted in 2014 on a sub-sample of AHS in all EAG States namely Bihar, Chhattisgarh, Jharkhand, Madhya Pradesh, Odisha, Rajasthan, Uttarakhand & Uttar Pradesh and Assam. The survey has collected information on nutritional status of women, children (1 month and above) and men, prevalence of anaemia among women, children (6 month and above) and men, prevalence of hypertension and abnormal fasting blood glucose among women and men 18 years and above and utilization of iodized salt in households.
We have used a number of libraries and tools:
- Python3 - Used for scripting our model. And many of its libraries were used:
- Scikit-Learn 0.19.1
- Pandas 0.22.0 - For Data Cleaning.
- NumPy 1.14.2 - For mathematical opearations.
- Matplotlib 2.2.2 - For visualisations.
- Seaborn 0.8.1 - For visualisations.
- HTML5 - Used for Markup.
- CSS3 - For styling the Markup.
- Bootstrap3 - For pre-prepared styling classes.
- JQuery - For JavaScript sending requests to the web server.
- Flask - A python library used for creating the web server.
- Chart.JS - For visualisations.
- To install the libraries required to execute the scripts and run the web-application go to the root folder and execute the following commands:
$ pip3 install -r requirements.txt- Generate the model.sav file by executing following command
$ python3 model.py- Go to the root directory of the project folder and execute the following commands:
$ python3 prediction.py- In a new terminal window execute the following command:
$ xdg-open visualisation/index.html