The following question has been closed because it is off-topic:
An approximation for path integral kernels
However, I'm presenting a new idea in that post: Set N=2 in the formula of the path integral kernel and derive an approximation for the kernel. This question is definitely not a homework and I'm not asking anyone to check the derivation. I have checked numerically that the derivation is correct.