2 ms·
Finding the global optimum is a lot harder than just finding a good local optima quickly. Which is good enough for ML.
by Houshalter 9y ago
Finding the global optimum is a lot harder than just finding a good local optima quickly. Which is good enough for ML.