PROACTIVE DISASTER DETECTION

PROACTIVE DISASTER DETECTION

Boddu Kushwanth Sai1, E Rahul2, Abhi C N3, Vadlamudi Narendra4

1Student in Computer Science and Engineering & Presidency University, Bengaluru

2 Student in Computer Science and Engineering & Presidency University, Bengaluru

 3 Student in Computer Science and Engineering & Presidency University, Bengaluru

4 Student in Computer Science and Engineering & Presidency University, Bengaluru

Abstract - This paper introduces the Disaster Prediction Dashboard, a machine learning tool designed to predict the risks associated with hurricanes, earthquakes, and floods. It allows users to upload their own datasets, effectively managing missing values through imputation and enhancing accuracy with scaling techniques. By utilizing models such as random forest and logistic regression, it forecasts hazardous events based on various environmental factors, including wind speed, temperature, rainfall, and soil moisture. The methodology encompasses data preprocessing, feature selection, model training, and accuracy evaluation. Users can easily interact with a straightforward interface to upload data, enter specific values, and receive forecasts accompanied by clear risk indicators. With a focus on achieving high accuracy in risk classification, the system aims to improve disaster preparedness and aid in decision-making. Its scalability and user-friendly design highlight the successful application of machine learning in tackling environmental challenges.

Key Words: Disaster prediction, machine learning, Random Forest, Logistic Regression, risk assessment