Contributed Poster Presentations: Business and Economic Statistics Section

Model fairness issues have received considerable attention in the ML community. The popular class of GLMs such as logistic regression has seen various developments to improve fairness. Despite the progress, studies remain limited in high-dimensional settings without adequate understanding from statistical perspectives. In this work, we propose a novel fairness-constrained model averaging algorithm for GLMs that can aggregate a large number of sparse model candidates to generate asymptotically fair modeling solutions. It is shown to achieve near optimal estimation risk in combining for estimation improvement, including flexible scenarios that a true model does not exist or is otherwise non-sparse. To facilitate practical applications for the model averaging approach, we further propose a new fairness-assisted stepwise sparsity learning method to help generate potentially fair model candidates. In addition, the fairness-assisted stepwise method with model selection maintains consistency properties when the true model is among the sparse feasible candidates, showing delicate distinction of combining for estimation improvement versus adaptation. 

This paper examines gold price dynamics using two complementary statistical approaches: the Geometric Random Walk (GRW) model and ARIMAX modeling, including ARIMA(0,1,1) with drift. The ARIMAX framework investigates the relationship between gold prices and global GDP, comparing log-level ARIMAX(1,0,1) and first-difference logarithmic models. Analysis reveals a statistically significant relationship, where a 1% increase in GDP corresponds to a proportional rise in gold prices. While the log-level model offers intuitive insights despite non-stationarity, both models highlight substantial uncertainty through wide prediction intervals. Forecasts are extended to 2030, demonstrating that all models predict rising gold prices but emphasize the challenges of long-term forecasting due to significant prediction intervals. The GRW model provides a robust benchmark for evaluating expert predictions and capturing uncertainty in future price movements. The addition of ARIMA(0,1,1) with drift captures trend-following behavior, enhancing understanding of gold price dynamics over time.
This integrated approach offers a comprehensive framework for gold price forecasting. 

This study examines the influence of microfinance institutions' (MFIs) financial innovation on structural transformation. For this purpose, we considered a household survey from Nepal. The survey collected data on various individual and household characteristics, borrowing patterns, and occupations over the years. The key question focused on occupations before and after borrowing, a categorical response variable indicating 1 for occupational change after borrowing and 0 otherwise. Therefore, we use logistic regression to estimate the probability of occupational change, given two measures of financial innovation: loan purpose and size. The results show that the number of households involved in agriculture significantly decreased, with the majority switching to businesses and convenience stores, indicating a shift to the manufacturing and service sectors. These findings suggest that MFIs contribute to local-level structural transformation by enabling borrowers to move away from traditional employment. This study has important implications for policymakers, development practitioners, and academics interested in promoting economic development through microfinancing in low-income area.