Enhancing IoT Network Performance Through Predictive Modelling with Machine Learning Regression

Abstract

IoT has grown rapidly, linking billions of devices and collecting massive amounts of data. Early IoT networks were governed by rule-based algorithms that adjusted performance using parameters. As networks advanced, these traditional methods struggled to keep up with IoT devices dynamic nature, causing inefficiencies and performance bottlenecks. Advanced predictive modelling with machine learning (ML) allows IoT networks to anticipate and respond to changing situations in real time. This research uses machine learning regression models to predict network behaviours, optimize resource allocation, and improve data transmission and device interactions in IoT networks. The research models data flow, network traffic, and device interactions using machine learning regression to forecast and improve IoT network performance. It predicts network conditions to improve operational efficiency and latency. Before machine learning, static rules, set thresholds, and human network adjustments were frequently insufficient to meet IoT devices dynamic behaviour. Traditional IoT network management systems cannot dynamically adapt to device behaviour, resulting in poor resource allocation, latency, and network congestion. These static methods frequently perform poorly with complex and high- volume IoT device data. Real-time, adaptive network management systems are needed as IoT devices grow. Machine learning algorithms can recognize data flow and device activity patterns to better forecast and control network performance. This research uses predictive modelling to improve IoT network efficiency and overcome traditional system limitations. Latency, bandwidth use, and device communication delays are predicted using machine learning regression models in the proposed system. Analysis of historical and real-time data allows the system to dynamically distribute resources, balance network traffic, and reduce congestion. Regression models predict network conditions to optimize operations and network performance. These models enable real-time, intelligent changes, enhancing IoT device connectivity and resource use.