ADDRESSING IMBALANCED NETWORK TRAFFIC IN INTRUSION DETECTION USING MODIFIED RANDOM FOREST AND KDD DATASET

The system employs machine learning techniques on the KDD dataset for intrusion detection improvements. The Load Data module, the focus of the analysis process, loads the dataset either from a remote site or from a local site. It guarantees that the data is organized and stored in memory well enough to facilitate manipulation. For optimum performance of machine learning, one of the prerequisites for feature selection and preprocessing quality is also the loading of data. The module for Data Preprocessing proves a major contributor to raw data quality and usability enhancement. In this process, duplicates are eliminated, missing values treated, and dataset inconsistency corrections performed. Equally important would be encoding categorical variables and normalizing data to fit into the ML models used in this research project. Thus, this crucial module standardizes the data and ensures the integrity of the dataset for further processing. Thus, it raises the performance of intrusion detection systems by increasing model capability and efficiency. A clean and structured base will allow machine-learning algorithms to detect patterns and anomalies. Hence, it provides an excellent and reliable means of intruder detection, identifying prospective threats accurately.