This repository inclludes a deep learning model for predicting the air pollution value for the next hour based on the given input. The dataset contains reports on the weather and the level of pollution each hour for five years at the US embassy in Beijing, China.
The data includes the date-time, the pollution called PM2.5 concentration, and the weather information including dew point, temperature, pressure, wind direction, wind speed and the cumulative number of hours of snow and rain. The complete feature list in the raw data is as follows:
- No: row number
- year: year of data in this row
- month: month of data in this row
- day: day of data in this row
- hour: hour of data in this row
- pm2.5: PM2.5 concentration
- DEWP: Dew Point
- TEMP: Temperature
- PRES: Pressure
- cbwd: Combined wind direction
- Iws: Cumulated wind speed
- Is: Cumulated hours of snow
- Ir: Cumulated hours of rain
We used this data and frame a forecasting problem where, given the weather conditions and pollution for prior hours, we forecast the pollution at the next hour.
The RMSE (Root Mean Square Error) value of the model calculated as 24 while the mean value of the pollution was 98. RMSE punishes large errors more, so while the model performed pretty good with the general of the data, it naturally struggled more with the outline data and this is the main factor that increases the RMSE. In conclution, we can safely say that the model performs pretty good in general.