C2_W1_Lab01_Neurons_and_Layers, I dont understand how the "w" parameter vectors results in [[0.08]] and "b" bias parameter to be 0

In cell 5 and 6, I dont understand why does giving the X_train[0], which is 1.0 would outputs 0.08 in the w parameter

As you can see in the following line, the weight a has been initialized to 0.08 and bias b to 0, so ax + b = 0.08*1 + 0.

The model needs to train to adjust its weights/bias better and giver a better output with time.

But how is it initialized to 0.08? is it a default initial parameter from keras.layers.Dense?

Hello, @Zolids,

0.8 was a randomly initialized parameter. The printed value is likely rounded to the first decimal place. The action of initialization was triggered by passing a sample through the layer.


1 Like