Compute the policy gradient by old fasion style, finite difference.
- chainer
- numpy
- matplotlib
[-0.565921, -0.2087926, -2.46556049, -0.91585176, 0.33338258, 0.41990558, -5.056393, 0.57013692, 1.08717598, -0.2268253, 0.28361697, -0.76006312, -1.92836859, -3.53624301, -2.75193499, 1.73638419]