Build a model to forecast future hospital emergency admissions based on past emergency admissions.
The data used in this project is obtained from the Hospital Episodes Statistics (HES) dataset which consists of the record of all patients admitted to NHS (public) hospitals in the United Kingdom.
Complete dataset available on the NHS Digital website
The modules needed to run this project can be installed using pip
-
Preprocessing the data:
- Normalize the data: to obtain better results
- Interpolate the data: to handle the missing values
- Split data into training set and test set
- Format the data: to transform the time series problem to a supervised learning problem
- Reshape data: give the data the 3D shape expected by LSTMs
-
Implement the model:
- Build the LSTM model
- Train the model
-
Make predictions:
- Use the model to make predictions
- Plot the predicted values compared to the actual values