Journal of Electrical and Computer Engineering

Research Article

A Hardware Design of Neuromolecular Network with Enhanced Evolvability: A Bioinspired Approach

Algorithm 1

Evolutionary learning algorithm.
(1) Generate at random the initial MAP, PU-Type, readin enzyme, and readout enzyme
patterns of each neuron in the reproduction subnet. Each neuron is denoted by
neuron ( 𝑠 , 𝑏 ) where 𝑠 is the subnet number and 𝑏 is the bundle number.
(2) Copy the MAP, PU-Type, readin enzyme, and readout enzyme patterns of each
neuron in the reproduction subnet to those of comparable neurons in the competition
subnets.
Copy neuron ( 5, 𝑏 ) to n e u r o n ( 1 , 𝑏 ) n e u r o n ( 2 , 𝑏 ) n e u r o n ( 3 , 𝑏 ) n e u r o n ( 4 , 𝑏 ) , for 𝑏 = 1 , 2 , … , 32
(3) Vary the MAP pattern of each neuron in the first subnet, the PU-type pattern in the
second subnet, the readin enzyme pattern in the third subnet, and the readout enzyme
pattern in the fourth subnet.
Vary t h e 𝑀 4 𝑃 p a t t e r n o f n e u r o n ( 1 , 𝑏 ) t h e 𝑃 𝑈 - 𝑇 𝑦 𝑝 𝑒 p a t t e r n o f n e u r o n ( 2 , 𝑏 ) t h e 𝑟 𝑒 𝑎 𝑑 𝑖 𝑛 𝑒 𝑛 𝑧 𝑦 𝑚 𝑒 p a t t e r n o f n e u r o n ( 3 , 𝑏 ) t h e 𝑟 𝑒 𝑎 𝑑 𝑜 𝑢 𝑡 𝑒 𝑛 𝑧 𝑦 𝑚 𝑒 p a t t e r n o f n e u r o n ( 4 , 𝑏 ) , if 𝑈 ≤ 𝑃 , f o r 𝑏 = 1 , 2 , … , 3 2
where 𝑃 is the mutation rate and 𝑈 is a random number generated between 0 and 1.
(4) Evaluate the performance of each competition subnet and select the best-performing
subnet.
(5) Copy the MAP, PU-Type, readin enzyme, and readout enzyme patterns of each neuron
in the best-performing subnet to those of comparable neurons in the reproduction
subnets, if the former shows better performance than the latter.
(6) Go to Step 2 unless the stopping criteria are satisfied.