Why is the probability predicted by my keras model always zero

Question

I\'ve built a model to predict loan suitability on a Kaggle dataset here

dataset = df.values
X = dataset[:,0:11].astype(float)
Y = dataset[:,11]
scaler = Stan