How to do incremental learning while building a model. Do we have a specific framework on this?
Ex: Let us say i have built a classification model on 50 GB of data. The model is good and we are using it. Now tomorrow if 1 gb new data comes in, how to incrementally train the model.i dont want to run the entire data to be passed again as it is computationally expensive . how to do this?
Any language/tool will do? i am not referring to scoring but adding incremental data and training the model in quick time.