I am a beginner at machine learning and Accord.net. I see how I can train it and then use it, but in an enterprise setting, running this on a server, how do I persist the learnings of Accord .net. Do I need to save all training data and retrain every time the application is restarted?
Ususally there are Load and Save methods for persistance and re-creation of state.
At which specific model are you looking at?
Import the Accord.IO namespace in your code, by adding
using Accord.IO;
at the top of your source file. This way, you will be able to use the Serializer's Save methods in any framework object. To load an object back from its serialized form, you can use Serializer.Load(string path).
来源:https://stackoverflow.com/questions/36889043/how-do-i-save-the-learned-state-in-accord-net