It may not be an "application of statistics" in the direct sense, but that depends on how one defines statistics, which brings one back to the original problem.

It may not be an "application of statistics" in the direct sense, but that depends on how one defines statistics, which brings one back to the original problem. Statistics and ML often have similar goals, but ML emphasizes computational efficiency over trace-able accuracy. Thus, I view each field as having different weights on the same sub-goals.