I'm very new to tensorflow and I'm trying to figure out some tricks for future implementation.
I have a dataset consisting of three column, a, b and c. I'm trying to create a model which can predict c from a and b. By construction of the data, it is known that c=b/a. I want to create a model which can capture the relation between the three columns s.t. it can be added to future models. This is what I have tried:
normalizer = tf.keras.layers.Normalization(axis=-1)
tfd = tfp.distributions
model=tf.keras.Sequential()
model.add(normalizer)
model.add(tf.keras.layers.Dense(units=10,activation='relu'))
model.add(tf.keras.layers.Dense(units=1))
Where the model has been fitted to a train set and tested on a seperate test set.
How can I create a model which can for example cross divide all columns with eachother and optimize through such a layer.