Scribd
Upload
76 views15 pages

MAT6007 Session4 MP Neuron Perceptrons

This document provides an overview of McCulloch-Pitts neurons and perceptrons for deep learning. It discusses how McCulloch-Pitts neurons use thresholds to classify inputs in a linearly separable manner. Perceptrons were proposed as a more generalized model that can learn weights and thresholds over time. The perceptron model assigns numerical weights to inputs rather than limiting them to boolean values. Perceptrons can implement linearly separable functions like the McCulloch-Pitts neuron but allow weights and thresholds to be learned from data.

Uploaded by

manojnaidu yandrapu
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PPTX, PDF, TXT or read online on Scribd
76 views15 pages

MAT6007 Session4 MP Neuron Perceptrons

This document provides an overview of McCulloch-Pitts neurons and perceptrons for deep learning. It discusses how McCulloch-Pitts neurons use thresholds to classify inputs in a linearly separable manner. Perceptrons were proposed as a more generalized model that can learn weights and thresholds over time. The perceptron model assigns numerical weights to inputs rather than limiting them to boolean values. Perceptrons can implement linearly separable functions like the McCulloch-Pitts neuron but allow weights and thresholds to be learned from data.

Uploaded by

manojnaidu yandrapu
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PPTX, PDF, TXT or read online on Scribd
You are on page 1/ 15

MAT6007 Deep Learning

McCulloch Pitts Neuron - Perceptron

Prakash P VIT , Chennai


Road Map
 McCulloch Pitts Neuron
 Perceptron

Prakash VIT Chennai


Geometric Interpretation of a MP Neuron
Already discussed that the OR function’s thresholding parameter theta is 1

Now plotting them on a 2D graph and making use of the OR function’s


aggregation equation

x1 + x2 ≥ 1 using which we can draw the


decision boundary as shown in the graph

Graphically show that all those inputs whose output when passed through the OR function M-P neuron lie ON or
ABOVE that line and all the input points that lie BELOW that line are going to output 0

The M-P neuron just learnt a linear decision boundary 


The M-P neuron is splitting the input sets into two classes — positive and negative. Positive ones (which output 1) are those that lie ON or
ABOVE the decision boundary and negative ones (which output 0) are those that lie BELOW the decision boundary
Geometric Interpretation of a MP Neuron
Geometric Interpretation of a MP Neuron

If you have more than 2 inputs?


Geometric Interpretation of a MP Neuron

Lets just generalize this by looking at a 3 input OR function M-P unit. In this case, the possible inputs are 8 points — (0,0,0),
(0,0,1), (0,1,0), (1,0,0), (1,0,1)……
Limitations of a MP Neuron
 What about non-boolean (say, real) inputs?

 Do we always need to hand code the threshold?

 Are all inputs equal? What if we want to assign more


importance to some inputs?

 What about functions which are not linearly separable?


 XOR function.
Classical Perceptron
Frank Rosenblatt, an American psychologist, proposed the classical perception
model, the mighty artificial neuron, in 1958

It is more generalized computational model than the McCulloch-Pitts neuron


where weights and thresholds can be learnt over time

A more general computational model than McCulloch–Pitts neurons

Main differences with MP Neuron:


 Introduction of numerical weights for inputs and a mechanism for learning
these weights
 Inputs are no longer limited to boolean values

Refined and carefully analysed by Minsky and Papert (1969) - their model is referred to as the perceptron model
Perceptron Model
Perceptron Model
Consider the task of predicting whether I
would watch a random game of football on
TV or not

Here, w0 is called the bias because it represents


the prior

A football freak may have a very low threshold The point is, the weights and the bias
and may watch any football game irrespective will depend on the data ( viewing
of the league, club or importance of the game history in this case)
[theta = 0]
Based on the data, if needed the model
A selective viewer like me may only watch a football may have to give a lot of importance
game that is a premier league game, featuring Man (high weight) to the isManUnitedPlaying
United game and is not friendly [theta = 2] input and penalize the weights of other
inputs
Perceptron vs McCulloch-Pitts Neuron
From the equations, it is clear
that even a perceptron separates
the input space into two halves,
positive and negative

All the inputs that produce an output 1 lie on one side (positive half space) and all the inputs
that produce an output 0 lie on the other side (negative half space)
A single perceptron can only be used to implement linearly separable functions, just like the
M-P neuron
Then what is the difference? Why do we claim that the The weights, including the threshold can be
perceptron is an updated version of an M-P neuron?
learned and the inputs can be real values
Boolean Functions Using Perceptron

One possible solution to this set of


inequalities is w0 = −1, w1 = 1.1, , w2 = 1.1

you can try it out for AND function


Errors and Error Surfaces
Let us fix the threshold (−w0 = 1) and try different values of w1, w2

Say, w1 = −1, w2 = −1 w1 = 1.5, w2 = 0 w1 = .45, w2 = 0.45

We are interested in finding an algorithm which finds the


values of w1, w2 which minimize this error  
References
 https://
towardsdatascience.com/mcculloch-pitts-model-5fdf65ac5dd1

 https://www.cse.iitm.ac.in/~miteshk/CS6910.html#schedule

Prakash VIT Chennai


Thanks

Prakash VIT Chennai

You might also like