In this Coding TensorFlow episode, Magnus gives us an overview of a common machine learning problem, overfitting and underfitting. The goal of preventing overfitting is to develop models that generalize well to testing data, especially data that they haven't seen before. Where as, in underfitting your model has room for improvement. It may not have learned the important patterns in the training data. Watch to learn more, follow along with the link below, and stay tuned for part 2 of this Coding TensorFlow episode!
Explore overfitting and underfitting written tutorial → bit.ly/2RFJR2O
Follow along in Colab! → bit.ly/2yAwubE
Watch our Text Classification videos:
Pt.1 → bit.ly/2OkYWJ9
Pt.2 → bit.ly/2pVwhMh
Learn more on overfitting and underfitting in this video → bit.ly/2Ptf1sR
Check out the Machine Learning Crash Course
with TensorFlow APIs → bit.ly/2MLUDkU
Watch more Coding TensorFlow → bit.ly/2zoZfvt
Subscribe to TensorFlow → bit.ly/TensorFlow1
Негізгі бет Ғылым және технология Solve your model’s overfitting and underfitting problems - Pt.1 (Coding TensorFlow)
Пікірлер: 22