In software defect prediction, noisy attributes and high-dimensional dataremain to be a critical challenge. This paper introduces a novel approachknown as Multi Correlation-Based Feature Selection (MCFS), which seeksto address these challenges. MCFS integrates two feature selectiontechniques, namely Correlation-Based Feature Selection (CFS) andCorrelation Matrix-Based Feature Selection (CMFS), intending to reducedata dimensionality and eliminate noisy attributes. To accomplish this, CFSand CMFS are applied independently to filter the datasets, and a weightedaverage of their outcomes is computed to determine the optimal featureselection. This approach not only reduces data dimensionality but alsomitigates the impact of noisy attributes. To further enhance predictiveperformance, this paper leverages the Particle Swarm Optimization (PSO)algorithm as a feature selection mechanism, specifically targetingimprovements in the Area Under the Curve (AUC). The evaluation of theproposed method is conducted on 12 benchmark datasets sourced from theNASA MDP corpus, renowned for their noisy attributes, highdimensionality, and imbalanced class records. The research findingsdemonstrate that MCFS outperforms CFS and CMFS, yielding an averageAUC value of 0.891, thereby emphasizing its efficacy in advancingclassification performance in the context of software defect prediction usingK-Nearest Neighbors (KNN) classification.