From Andrew Ng ML course:
Simplifying the classification problem on the right as a binary \(0/1\) \(x_1\) and \(x_2\) prediction we have the equivalent of an XOR gate:
In fact, if the circles are zeros, the model is more of an XNOR function: \(y = x_1 \text{XNOR} x_2 = \text{NOT} (x_1 \text{XOR} x_2)\).
NOTE: These are tentative notes on different topics for personal use - expect mistakes and misunderstandings.