Abstract
We propose an active learning strategy for regression-model-based QoT estimation to achieve higher accuracy with fewer samples. Results show that our strategy decreases 35.8% MSE on average than traditional method with the restricted sample size.
© 2021 The Author(s)
PDF Article