Freigeben über


Supervised Machine Learning Part 3- Benchmarking

Now that we have extracted our dataset and have chosen which supervised ML models to classify our data with, it is time to benchmark and optimize our model. A common mistake is to train and test a classification model using all the documents in your data set. If you do this your model will super-fit […]