Splitting and Parallelizing of Quantum Convolutional Neural Networks for Learning Translationally Symmetric Data