May 2018
Beginner
490 pages
13h 16m
English
A linear model cannot solve the XOR problem expressed as follows in a table:
| Value of x1 | Value of x2 | Output |
| 1 | 1 | 0 |
| 0 | 0 | 0 |
| 1 | 0 | 1 |
| 0 | 1 | 1 |
The following graph shows the linear inseparability of the XOR function represented by one perceptron:

The values of the table represent the Cartesian coordinates in this graph. The circle with a cross at (1,1) and (0,0) cannot be separated from the circles at (1,0) and (0,11). That's a huge problem. It means that Frank Rosenblatt's f(x,w) perceptron cannot separate, and thus not classify, these dots into clouds and trees; an object ...
Read now
Unlock full access