Cybersecurity Insights: Analyzing IoT Data Through Statistical and Visualization Techniques

The rising frequency of cyberattacks on Internet of Things (IoT) networks calls for sophisticated approaches to threat analysis and mitigation. The importance of data analysis-especially visualization-in identifying patterns that can guide successful mitigation efforts is emphasized in this research. To address class imbalance issues, we first developed thorough data preparation and basic class analysis. Next, we used the BoT-IoT and TON-IoT datasets to visualize an analysis of 5-tuple network properties. The Spearman and Pearson correlation coefficients were used to statistically validate these properties. Furthermore, class overlapping was revealed using unsupervised techniques using PCA and T-SNE, which shed light on the characteristics of IoT data. Our results show how these visualization methods can help reveal patterns in IoT datasets and illuminate the nature of the representative IoT data, in addition, this approach contributes to the development of robust cybersecurity safeguards for IoT security in the future.