Optimization algorithms
Optimization is the key to how a network learns. Learning is basically an optimization process. It refers to the process that minimizes the error, cost, or finds the locus of least errors. It then adjusts the network coefficients step by step. A very basic optimization approach is the one we used in the previous section on gradient descents. However, there are multiple variations that do a similar job but with a bit of improvement added. TensorFlow provides multiple options for you to choose as the optimizer, for example, GradientDescentOptimizer, AdagradOptimizer, MomentumOptimizer, AdamOptimizer, FtrlOptimizer, and RMSPropOptimizer. For the API and how to use them, please see this page:
https://www.tensorflow.org/versions/master/api_docs/python/tf/train#optimizers ...
Become an O’Reilly member and get unlimited access to this title plus top books and audiobooks from O’Reilly and nearly 200 top publishers, thousands of courses curated by job role, 150+ live events each month,
and much more.
Read now
Unlock full access