McCulloch-Pitts Neuron it is known as Mankinds first Algorithm for deep learning. It is based on normal mathematical functions/operations to train the program. It is basically based on neurons of Human brains.
Dendrite: Receives signals from other neurons
Soma: Processes the information
Axon: Transmits the output of this neuron
Synapse: Point of connection to other neurons
McCulloch-Pitts Neuron
It uses binary activation function which means the inputs should be only in 2 different numbers 0 or 1 and the algorithm will give results in the form of 0 and 1 only.
- The firing state is binary (1= fire, 0=not fire).
- Weights in the network remain the same throughout the program.
- It was able to solve Linear separable problems.
- There is no bias. This algorithm is known as the unbiased deep learning algorithm.
- It was able to solve Logical Gates AND, OR, NOT.
AND GATE:
if(sum)>threshold:
output=1
else:
output=0
Steps to solve manually:
- Assume weight = 0(discard because weight cannot be 0 if we multiply it with any given input it will result in 0 itself).
- Assume weight = 1.
- Using the formula we will calculate threshold value using the input x1,x2 from the input of AND gate.
x1w+x2w= 1x1+0x1- From step 4 we can get the value 2 which is value for the threshold.
NAND GATE
Let’s solve NAND Gate using the same steps above:
Let’s assume w=-1 so when x1=x2=0.
yin=x1w1+x2w
(0x-1 + 0x-1=0)
when
x1=1
x1=1 =>1x-1+0x-1= -1
x2=0
x1=0 0x-1+1x-1=-1
x2=1
x1=11x-1+1x-1=-2
x2=1
Result:
0 1 1
-1 1 1
-1 1 1
-2 0 0
the algo:
if yin>=-1 then y=1 else: y=0
Your task is to solve OR, NOT Gate using the steps above.
