MATLAB: What learning rules can be used for incremental training in the Neural Network Toolbox

Deep Learning ToolboxincrementallearningNetworkneuraltraining

I would like to use the ADAPT function to do incremental training. Can I use TRAINLM for weight and bias learning or am I restricted to LEARNGD and LEARNGDM?
What choices do I have for the BTF, BLF, and PF fields in the newff statment if I want to use adapt to train?

Best Answer

The incremental learning rules LEARNGD and LEARNGDM are the only incremental training algorithms in the Neural Network Toolbox. TRAINLM is not an incremental learning rule because, like other non-incremental algorithms such as TRAINGDX, TRAINGDX etc, TRAINLM requires all the vectors to be present for it to work.
If you are using ADAPT to train, BLF can be LEARNGD or LEARNGDM. PF can be MSE, MSEREG, or SSE. BTF is not used when using ADAPT to perform incremental training.
Related Question