<![CDATA[Abstract Generation Z's rapid uptake of financial technology (fintech) presents both an opportunity for financial inclusion and a challenge for providers seeking to understand the drivers of adoption. While prior studies have relied predominantly on covariance or variance based structural equation modeling (SEM) to explain fintech adoption intention, machine learning (ML) classification algorithms offer a complementary, prediction-oriented approach capable of capturing non-linear relationships among adoption determinants. This study compares eight supervised classification algorithms Logistic Regression, Decision Tree, Random Forest, K-Nearest Neighbors, Support Vector Machine, Naïve Bayes, a Multi layer Perceptron Neural Network, and XGBoost in predicting fintech adoption intention among Generation Z users, using predictors derived from the Unified Theory of Acceptance and Use of Technology 2 (UTAUT2) framework (performance expectancy, effort expectancy, social influence, facilitating conditions, hedonic motivation, price value, habit) augmented with perceived risk and financial literacy. Using dataset a survey of 300 Generation Z respondents model performance was evaluated using accuracy, precision, recall, F1-score, ROC-AUC, and 10-fold cross validation. In the illustrative run, the Neural Network and XGBoost models achieved the most favorable discrimination (AUC ≈ 0.57–0.61), while effort expectancy, facilitating conditions, social influence, and habit emerged as the most important predictors based on Random Forest feature importance. These findings, once confirmed with primary survey data, are expected to provide fintech providers and policymakers with a prediction oriented complement to traditional SEM-based adoption studies, supporting more targeted onboarding and engagement strategies for Generation Z users. Keywords: fintech adoption; Generation Z; machine learning; classification algorithms; comparative analysis.]]>
Copyrights © 2026
