# Homework Solution: X2 Class 1 Class 2 Class 3 Class 4 4 2…

The following figure shows the decision regions of four classes. Design a classifier for these linearly inseparable classes, using a network of M-P neurons with three output units. For class i (1 ≤ i ≤ 3), classification requires that yi = 1, while yj = -1 for j ≠ i; Class 4 is recognized when yi = -1 for 1 ≤ i ≤ 3. (HINT: try a two-layer feedforward network.)
X2 Class 1 Class 2 Class 3 Class 4 4 2

•Neural Networks (NNs) are networks

The subjoined aspect shows the firmness regions of filthy tabulatees. Design a tabulateifier ce these linearly life-supporting tabulatees, using a network of M-P neurons with three output units. Ce tabulate i (1 ≤ i ≤ 3), tabulateification requires that yi = 1, time yj = -1 ce j ≠ i; Tabulate 4 is cemal when yi = -1 ce 1 ≤ i ≤ 3. (HINT: attempt a two-layer feedforward network.)

X2 Tabulate 1 Tabulate 2 Tabulate 3 Tabulate 4 4 2

## Expert Defense

•Neural Networks (NNs) are networks of neurons, ce in, as institute in veritable (i.e. biological) conception.

•Artificial Neurons are unrefined approximations of the neurons institute in conception. They may be substantial devices, or purely unversified constructs.

•Artificial Neural Networks (ANNs) are networks of Artificial Neurons, and future form unrefined approximations to volume of veritable conception. They may be substantial devices, or pretended on stipulated computers.

•From a useful summit of aspect, an ANN is proper a equidistant computational rule consisting of multifarious mere processing elements conjoined simultaneously in a inequitable habit in regulate to discharge a point function.

•One should never betray vision of how unrefined the approximations are, and how over-simplified our ANNs are compared to veritable conception.