Earthquake Alert

Inspiration

Changes in global climate change have given rise to subsequent natural crises, this has caused a lot of damage to humans and property. in this project, I decided to build a machine learning model trained on previous earth-quake occurrences to predict future occurrences in the next 5 days

What it does

This model is going to use a set of data to train and learn patterns so it can predict earth-quake occurrence in future. Earthquake occurs at 1 and donot occur at 0

How we built it

It was built using machine learning algorithms and a supervised learning approach and uses real time data that updates every minute on https://earthquake.usgs.gov/earthquakes/feed/v1.0/csv.php for past 30 days. I have used gridsearch CV for improving model and hyperparameter tunning on DecisionTreeClassifier and RandomForestClassifier

• Using the same hyper parameters I trained XGBoost. ## Challenges we ran into The very challange we runned into was in the inference stage where I had to build an python web app to receive real-time data and process using flask ## Accomplishments that we're proud of With the use of xgboost the model is able to predict with the AUC score of an accuracy of 0.98 ## What we learned I learnt to build a robust pipeline to with-stand future data entries ## What's next for Earth Quake Alert I aim to build an Andriod and IOS app using tensorflow lite, Andriod Studio and flutter

Built With

• flask
• numpy
• python
• sklearn
• xgboost