Method for Profile-Guided Optimization of Android Applications Using Random Forest

When choosing a smartphone, many users are guided by the performance of smartphones and speed applications. Because it is difficult to measure application directly, startup considered used for comparison. Android runtime (ART) uses several technologies up The first approach just-in-time (JIT) compilation frequently methods at runtime. second compile entire code ahead time (AOT). profile way strike balance between JIT AOT. optimization features were introduced in Nougat form profile-guided (PGO). By aggregating data from multiplicity devices Play Store, ART profiling significantly speeds this process makes its outcome available all alike. We propose machine learning based method called SPMLGen generating profiles optimization. avoid delays caused need collect information advance perform ensure user privacy. With generated SPMLGen, we obtain approximately same launch as with Store. Measurements taken on Samsung Galaxy S22 A52 12 firmware dozen