An exploratory earthquake analysis of earthquakes in Turkey. And use of machine learning algorithms to predict the future earthquakes of Istanbul. Weather data is tried to be associated with the earthquake data.
- with python
- libraries used: scikit, matplotlib, seaborn, numpy, pandas.
This project is made for the Data Mining Course that I was taking during my bachelor's degree at Işık University.
I also have written an article (unpublished) with the help of my instructor. You can read it here. It involves the methods used, experiments conducted and the results obtained throughout this study.
- I used Bogazici University Kandilli Observatory and Earthquake Research Institute data for earthquakes in Turkey.
- There is a .txt formatted version of the data above here. This dataset represents earthquakes in 1910-2017 where intensity is greater than 3.5.
- I also used Meteorological Service data to observe past weather data in 1927-2019.
- Where does earthquake occur in Turkey the most in terms of city and region?
- In which season does earthquake occur the most?
- How much Turkey’s weather is out of its normal values?
- Which area of Turkey is the most effected from climate change?
- My hypothesis is that as climate distorts earthquakes will increase.
- I calculated the frequency of earthquakes in cities and regions of Turkey.
- I drew graphs to indicate earthquake occurrence per year.
- I compared expected (normal) weather and realized weather in each city in yearly basis.
- I drew graphs to indicate climate distortion.
- I looked for a correlation between increase of earthquake and climate change.
- I tried to predict future earthquakes.