Institute of Technology Tralee | Big Data
Dataset: Climate Change: Earth Surface Temperature Data
Time Series Prediction Algorithm: ARIMA model
Conclusion:
Climate change remains as a global issue and the global land and ocean average temperature is gradually increasing. ARIMA time series prediction model offers an insight for planning the suitable climate change action and strategy and aids to bring climate action to top of the international agenda. The result of the forecast validation yields an MSE of 0.01032 and an RMSE of 0.1016. It was concluded that the global land and ocean average temperature from 2016 to 2020 is forecasted to have seasonal fluctuated temperature. However, it was expected to have a gradually increasing predicted temperature due to an increase in emission of CO2. Seasonality of the selected dataset for time series prediction can be examined and removed that causes data to be non-stationary which violates the assumptions of the ARIMA model before executing time series prediction in future work.
NOTE: More information is available on attached research report.