Anoise / Minusformer

Improving Time Series Forecasting by Progressively Learning Residuals
50 stars 11 forks source link

Why do you use the weighted subtraction rather than weighted addition? #2

Open erikalien5595 opened 1 month ago

erikalien5595 commented 1 month ago

Dear Author, I have greatly appreciated your insightful work, and I would like to seek clarification on a particular point that has been troubling me. In Section 2.2 of your paper, you state that "The overall estimation $\mathcal{𝐹}(𝑋)$ in the deep model is a weighted subtraction of the $L$ estimations." I am curious as to why you chose to employ weighted subtraction instead of weighted addition. I would be grateful for your explanation on this matter. Thank you very much for your time and consideration.