Software aging is the major cause of system failure as the software system runs consistently for a longer duration of time, which degrades the quality of the software. Aging-Related Bugs (ARBs) are caused due to a null pointer exception, memory leakage, unreleased locks, etc. Thus, timely prediction of such bugs is necessary for saving the software system from failure. Dynamic Ensemble Selection (DES) is a way to select the most compatible base classifiers from the pool of classifiers for a specific instance. In this paper, we empirically validated five DES algorithms for predicting the ARBs in the software system. The results are proven on seven open-source software systems falling broadly under two popular software, MySQL and Linux. The authors promote the use of DES algorithms for early prediction of ARBs in software. K-Nearest Output Profile (KNOP), one of the investigated DES algorithms outperformed the others by depicting the highest average F1-score, Matthews Correlation Coefficient (MCC) and Geometric Mean (G-mean) over seven datasets.