CS5720 - Week 1
Slide 8 of 20

Perceptron Limitations: XOR Problem

XOR Gate Truth Table

The XOR (exclusive OR) gate outputs 1 when inputs are different. This simple pattern exposed a fundamental limitation of perceptrons.
Input 1 (x₁) Input 2 (x₂) Output (y)
0 0 0
0 1 1
1 0 1
1 1 0
⚠️ The Problem
No single line can separate the 0s from the 1s in this pattern. This is what we mean by "not linearly separable."

Why Perceptrons Fail

The Mathematical Impossibility:

• A perceptron can only draw straight lines
• XOR requires curved boundaries or multiple lines
• No weights exist that satisfy all four cases
• This limitation stalled AI research for years!
Historical Impact:
Minsky and Papert's 1969 book highlighted this limitation, causing the first "AI Winter" - a decade-long pause in neural network research.

XOR Pattern Visualization

x₂
x₁
0
1
1
0

Classification Challenge

Try to imagine drawing a single straight line that puts all red circles (0s) on one side and all green circles (1s) on the other. It's impossible!

🚫 Single Perceptron Cannot Solve XOR
This discovery was devastating but also enlightening. It showed that we needed multi-layer networks (hidden layers) to solve non-linearly separable problems. The solution? Stack perceptrons in layers!
Prepared by Dr. Gorkem Kar