search

foryichuanqi / RESS-Paper-2022.09-Remaining-useful-life-prediction-by-TaFCN

The source code of paper: Trend attention fully convolutional network for remaining useful life estimation in the turbofan engine PHM of CMAPSS dataset. Signal selection, Attention mechanism, and Interpretability of deep learning are explored.
49 stars 2 forks source link

How to calculate lossb ？ #2

Closed Lunarli closed 2 years ago

Lunarli commented 2 years ago

In your paper you said ： image and x was an independent variable But in you sharing code： image The lossb was only related to the ytrue label Did I misunderstand? What exactly does x mean?

foryichuanqi commented 2 years ago

Exactly. The lossb was only related to the ytrue label. lossb get the minimum value of the ( ) of Eq.(1) when x=-0.5 A. -0.5 A is the average value of ytrue label. Why can lossb get the minimum value of the ( ) of Eq.(1) when x=-0.5 A? Because x=-0.5 A is obtained by taking the derivative of ( ) of Eq.(1)

Lunarli commented 2 years ago

Thanks! I got it! I'm just wondering what x stood for when I saw Eq.(1) in your paper. Actually，The lossb was the standard deviation of ytrue and independent of min function. BTW，your sharing code is pretty well. It has a complete structure and is very kind to reproduce. Thanks for sharing it !

------------------ 原始邮件 ------------------ 发件人: "foryichuanqi/Paper2022-Remaining-useful-life-prediction-by-TaFCN" @.>; 发送时间: 2022年11月25日(星期五) 中午11:21 @.>; @.**@.>; 主题: Re: [foryichuanqi/Paper2022-Remaining-useful-life-prediction-by-TaFCN] How to calculate lossb ？ (Issue #2)

Exactly. The lossb was only related to the ytrue label. get min value when x=0.5*A.

— Reply to this email directly, view it on GitHub, or unsubscribe. You are receiving this because you authored the thread.Message ID: @.***>

foryichuanqi commented 2 years ago

you're welcome