Using machine learning to find cloud?base height: a didactic challenge

This is an example of the training data used to learn how predict lowest cloud-base height using a neural network. The Illustration shows 1300 columns that form part array. Each column consists 280 rows, comprising 70 rows each standardised temperature, humidity and pressure indicating location cloud base in binary format. To benefit those wishing begin exploring machine learning atmospheric science context, we provide dataset some code, which can be train network given profiles pressure.