Bima Setyo Nugroho, Hujjatullah Fazlurrahman, Tony Seno Aji, Achmad Fitro, Maryam Jamilah Asha'ari, Anik Lestari Andjarwati, Audika Briantari Zaen, Ina Uswatun Nihaya
This study investigates how digital creativity and AI involvement predict graphic designers' performance using explainable machine learning approaches. A dataset of 352 labeled responses from Southeast Asian designers was analyzed, with performance measured as a composite score derived from client satisfaction, revenue growth, and project delivery consistency. After categorical encoding and scale normalization, five regression-based machine learning algorithms were tested under five-fold cross-validation. Random Forest achieved the best results (R2 = 0.465, with the lowest MSE), and SHAP analysis identified 'Human-AI Collaboration' as the most influential predictor. Optimal AI usage, defined as daily integration of diverse generative tools with high trust levels, was associated with higher creative output. The novelty of this research lies in integrating digital creativity and AI engagement indicators into an explainable predictive framework, complemented by clustering analysis and scenario-based simulations. However, generalizability is limited by the regional scope of respondents and the dominance of image-based AI tools, warranting further validation in broader contexts. © 2025 IEEE.
Universitas Negeri Surabaya, Faculty of Economics and Business, Department of Management, Surabaya, Indonesia; Universitas Negeri Surabaya, Faculty of Economics and Business, Department of Digital Business, Surabaya, Indonesia; Universitas Negeri Surabaya, Faculty of Economics and Business, Department of Economics, Surabaya, Indonesia; Graduate School of Business, University Kebangsaan Malaysia, Faculty of Economics and Management, Kuala Lumpur, Malaysia
