Tanmay Damle- Dept of computer Engineering VESIT, firstname.lastname@example.org
Harsh Deshmukh- Dept of computer Engineering VESIT, email@example.com
Anish Nair- Dept of computer Engineering VESIT, firstname.lastname@example.org
Digvijay Kocharekar- Dept of computer Engineering VESIT, email@example.com
Dr. Mrs.Sujata Khedkar- Dept of computer Engineering, VESIT, firstname.lastname@example.org
The societal impacts of flash floods are more significant than any other weather-related hazard. Flash floods are often manifested in the form of infrastructure damage, flooding roadways and bridges, creating deadly hazards to motorists and inundation of crops and pasture. For instance, the flash floods in Mumbai that occurred on 26th July 2005 caused a heavy amount of damage to human life and property.This system aims to predict flash floods in various parts of our country by learning from past data. The user will input parameters such as location, expected rainfall, water levels in nearby reservoirs etc. and the system will give the probability of flash floods in the region in the upcoming few days. A number of various ML models namely LogisticRegression,RandomForestClassifier,KNeighborsClassifier,VotingClassifier, GradientBoost are used and their performance is evaluated. Evaluation measures such as accuracy, recall, f1 score are used to validate the predictions. Based on these evaluation metrics the proposed system works most efficiently for GradientBoostingClassifier Algorithm with an accuracy of about 91%.