So an alternative to deep learning is tree based methods and gradient boosted methods on top of those trees. XGBoost etc. These aren't technically deep learning but they have a ton in common.