In this project I will demonstrate how to use Deep Learning Algorithms, such as the Recurrent [Deep] Neural Network in this example, to use SSN (Sunspot Number) to predict climate change.
"A measure of natural phenomenon, the sunspot number (SSN) is a solar activity index with long data record. It is frequently used when studying long-term phenomena like climate change, though it may not be the most appropriate index (Georgieva and Kirov, 2006)."
Simply clone the repository and run the .ipynb file in jupyter notebook, Goodgle Colabotary, or any other text editor that supports the iPython Notebook. And Just click the 'Restart & Run All' option and the file will run itself...
The Deep Neural Network model achieved the following results:
Train Score: 5.28 RMSE
Test Score: 5.37 RMSE
This implies that the model achieved around 95% accuracy in its prediction. You will see the close-proximity in the resemblance between the actual SSN graph and the predicted SNN graph.