Let's now see how a total of 201 parameters are obtained for this model. The dense_1 layer shows 140 parameters. These parameters are based on there being 13 units in the input layer that connect with each of the 10 units in the first hidden layer, meaning that there are 130 parameters (13 x 10). The remaining 10 parameters come from the bias term for each of the 10 units in the first hidden layer. Similarly, 50 parameters (10 x 5) are from the connections between two hidden layers and the remaining 5 parameters come from the bias term from each of the 5 units in the second hidden layer. Finally, dense_3 has 6 parameters ((5 x 1) + 1). Thus, in all, there are 201 parameters based on the architecture of the neural network model that was chosen in this example.