Optimization of computational performance in scientific Python — NumPy, Numba, Numexpr, Cython, and MPI — applied to NLP vectorization (TF-IDF, Naive Bayes) and Mandelbrot set calculation.