
S2 (Ep1) Down to Earth: A ”Time Series” of Machine Learning in Earth Observation
Down To Earth: A podcast for Geoscientists by Geoscientist
00:00
How Did the Back Propagation Algorithm Come to Be?
Back propagation is a fine tuning algaritm that was quick and easy to prograsm. It was ground breaking because it improved the accuracy neural network outputs, making them more useful. In the old days, they were very, very simple and were not very useful for classifying non linear data. We could not only learn linear to classify linearly, like we did in the neonetor prior to that. But e that was not the final solution. And i can tell you, when the back propagation algorithm came it was sort of the solution.
Transcript
Play full episode