Filter-based feature selection approaches are less computationally intensive and better suited for high-dimensional data. Given the many computational properties of multiple filter algorithms, determining the ideal one for the most outstanding performance is a complex problem. As a result, developing a suitable aggregation mechanism is critical to achieving optimal performance. Aggregating rankings often aims to find a consensus by minimizing the average difference between the combined ranking and individual ones. However, this approach can be unfair as the final ranking might favor some inputs over others, creating bias. This study introduces a rank aggregation methods that utilizes particle swarm optimization (PSO) as the aggregator. The objective is to minimize the distance between the input rank list and the output rank. The experiment was carried out on two publicly available biomedical datasets with varying numbers of features. Using three classifiers, the performance of the proposed approach was observed, and its superiority was shown in terms of both the number of features chosen and accuracy.