I'm using the Holt-Winters model in MATLAB to forecast weather data. When I set the smoothing parameters (Alpha, Beta, and Gamma) too high, the model overfits the data and doesn't forecast well on new data.
How can I adjust these parameters to avoid overfitting and improve the model's accuracy?
I tried setting the smoothing parameters (Alpha, Beta, and Gamma) to higher values in my Holt-Winters model, expecting it to make the model more responsive to recent data. However, this caused the model to overfit the training data, leading to poor forecast accuracy when predicting new data. The model became too sensitive to past data, failing to generalize well to unseen data.
发布者：admin，转转请注明出处：http://www.yc00.com/questions/1744351287a4569991.html
评论列表（0条）