A composite intelligence scoring framework for identifying high-potential individuals using multi-metric predictive models

The identification of intelligent individuals through culturally pertinent and objective evaluation frameworks is essential for the development of talent and the advancement of education. This study introduces a novel composite intelligence evaluation system that is specifically tailored to the socio-cultural and educational environment of Kazakhstan. The framework encompasses three critical domains: educational achievement, cognitive capabilities, and inventive performance. The study introduces the predictive intelligence analysis model (PIAM) and the dynamic intelligence scoring algorithm (DISA) to evaluate and predict high-potential individuals. A hierarchical weighted multi-metric integration model (HWMMIM) is employed in the methodology to evaluate the efficacy of innovation. This model incorporates sophisticated mathematical formulations, such as polynomial weighted GPA, harmonic mean-based cognitive indexes, and a recursive aggregation model. The DISA model obtained an AUC-ROC of 0.95, precision of 91%, recall of 89%, and accuracy of 94% on a dataset consisting of 10,000 individuals.. The composite intelligence score (CIS) is modified through logistic transformation to facilitate the probabilistic interpretation of classification problems. The proposed models facilitate strategic initiatives such as "Kazakhstan 2050" by enabling the identification of intellectual talent through the use of scalable, data-driven methodologies. In comparison to conventional IQ-based approaches, this research not only demonstrates improved prediction efficacy but also establishes a reproducible framework for culturally adaptive intelligence modeling in developing countries.