In the user manual, it shows the different kernel_initializer below
https://keras.io/initializers/
the main purpose is to initialize the weight matrix in the neura
GlorotUniform, keras uses Glorot initialization with a uniform distribution.r = √(3/fan_avg)
fan_avg = (fan_in + fan_out) /2
number of inputs = fan_in
number of nurons in a layer = fan_out