mekka-22 / flood-prediction-unit-hydrography-and-ml Goto Github PK

• This jupyter notebook uses 3 Machine Learning Algorithms namely KNN Classification, Logistic Regression[LR], Support Vector[SVM] to get the best possible model to predict the floods using Kerala Rainfall Data.

The state of Kerala does not experience floods as worse as the Indo-Gangetic Plains do but it is becoming more prone to flooding by the year. While Kerala floods largely occur as a result of incessant heavy rainfall, other factors that contribute to Kerala floods include mismanagement of water resources and forests. Human interventions including reclamation of wetlands and water bodies, construction and expansion of roadways, establishment of more and more settlements, deforestation, etc. have increased over the last few years in Kerala. It is estimated that 26% of the total geographical area of Kerala is prone to floods.

• Non-recording Gauges
• Recording Gauges
• Radar Measurement of Rainfall
• Observations by Satellite

• Pandas (for data representation)
• Numpy (for numerical computation)
• Matplotlib (for data visualization)
• Scikit Learn (for model building, training and testing)
• Seaborn (for data visualization)
• Django(to deploy the model)