Information-Centric Networking (ICN) is an architecture of the future internet that revolves around a shift in focus from host-based addressing to content-based retrieval. This paradigm shift brings about in-network caching and name-based routing into focus. However, traditional cache replacement strategies like LRU and LFU fail to adapt to changing request patterns and content popularity in real-time. In this work, we propose a dynamic selection of cache policies through the assistance of machine learning (ML) classifiers that analyze real-time network parameters to predict the optimal caching policy. On a Python-based simulation platform, we compare various ML models such as Decision Tree, Random Forest, K-Nearest Neighbors, and Logistic Regression. Results show improved performance compared to the conventional approach on cache hit ratio, latency, and hop reduction.