Junction of Quantum-Classical Computer-Aided Design Lab (JQub)

QuantumFlow Demo (1) --- Classify 3 and 6 in MNIST

QF-Net:

Layer 1:

Neural 1:
[1.,  1.,  1.,  1.,  1.,  1.,  1., -1.,  1., -1.,  1., -1.,  1.,  1.,	1.,  1.]
Neural 2:
[-1., -1., -1., -1., -1., -1., -1., -1., -1.,  1., -1.,  1., -1., -1.,-1., -1.]

Layer 2:

Neural 1:
[ 1., -1.]
Neural 2:
[-1., -1.]

BN for Layer 2:

Neural 1:
lt 0.5; 0.3060
Neural 2:
gt 0.5: 0.6940

Accuracy:

95.73% for MNIST with input downsampled to 4*4

QF-Circ:

QF-Circ Implementation in Quirk, based on the trained model of QF-Net

[Quirk link of the circuit without inputs](https://algassert.com/quirk#circuit=