Unveiling Patterns in Local Competitiveness: An Unsupervised Feature Selection Approach for Philippine Cities and Municipalities

While local competitiveness is an essential measure in evaluating localities’ performance, it is cumbersome due to the inclusion of numerous indicators. This study proposes an unsupervised feature selection method based on multiple criteria decision making (MCDM) and spectral clustering techniques to evaluate local competitiveness in Philippine cities and municipalities. The proposed method combines the PROMETHEE-GAIA algorithm with spectral clustering, which helps to reduce subjectivity and effectively handle complex data structures. The proposed algorithm is applied to analyze local competitiveness in Philippine cities and municipalities. The findings of this research reveal ten indicators that best represent the competitiveness of cities and municipalities in the Philippines. Moreover, most of these selected indicators reflect the municipality or city’s ability to support its local businesses. The findings of the study can be used to develop data-driven strategies for inclusive growth that will allow localities to thrive in a competitive environment. It simplifies the evaluation and provides clearer insight into the city’s job creation and investment potential.