Flood Forecasting Model Using Federated Learning

Abstract

The increasing sophistication of web attacks such
as SQL injection, Cross-Site Scripting (XSS), and
Distributed Denial of Service (DDoS) has elevated
the need for intelligent and adaptive detection
mechanisms. Traditional rule-based systems are no
longer sufficient due to their inability to detect
novel and evolving threats. This research aims to
investigate and compare the effectiveness of
different Machine Learning (ML) and Deep
Learning (DL) algorithms in detecting web attacks.
Algorithms such as Support Vector Machines
(SVM), Decision Trees, Random Forests,
Convolutional Neural Networks (CNNs), and
Recurrent Neural Networks (RNNs) are explored
and benchmarked. The study uses widely
recognized datasets like CICIDS2017 and UNSWNB15
to evaluate the models under consistent
conditions. Metrics such as accuracy, precision,
recall, and F1-score are used to gauge the
performance of each algorithm. The results indicate
significant differences in how well each model
detects certain types of attacks, and suggest the
potential of ensemble and hybrid models for realworld
web security systems. The goal is to
contribute toward more robust and scalable
intrusion detection solutions.