GLOSSARY TERM
What is Stochastic Gradient Descent?
An optimization method calculating the gradient using a single, randomly selected example.
SGD introduces high variance into parameter updates, which often allows the optimizer to escape shallow local minima. Mini-batch SGD strikes the optimal balance between computational efficiency and convergence stability.
Efficient Optimization
Tune the core optimization algorithms of your network via TacTech MLOps.