[1909.09020] Shape and Time Distortion Loss for Training Deep Time Series Forecasting Models

This paper addresses the problem of time series forecasting for non-stationary signals and multiple future steps prediction. To handle this challenging task, we introduce the Shape and Time Distortion Loss (STDL), a new objective function dedicated to training deep neural networks. STDL aims at accurately predicting sudden changes, and explicitly incorporates two terms supporting precise shape and temporal change detection. We introduce a differentiable loss function suitable for training deep neural nets, and provide a custom back-prop implementation for speeding up optimization. We also introduce a variant of STDL, which provides a smooth generalization of temporally-constrained Dynamic Time Warping (DTW). Experiments carried out on various non-stationary datasets reveal the very good behaviour of STDL compared to models trained with the standard Mean Squared Error (MSE) loss function, and also to DTW and variants. STDL is also agnostic to the choice of the model, and we highlight it

2 mentions: @evolvingstuff@vleguen
Date: 2019/09/20 02:18

Referring Tweets

@evolvingstuff Shape and Time Distortion Loss for Training Deep Time Series Forecasting Models paper: t.co/jgRT3tC5Df code: t.co/NIC6XjXlyn t.co/dopIL5qYYu
@vleguen Our paper "Shape and Time Distortion Loss for Training Deep Time Series Forecasting Models" accepted for NeurIPS 2019 conf now on Arxiv ! See you in Vancouver #NeurIPS #NeurIPS2019 t.co/dTBgl1XyCE t.co/aeqUArmfSq

Related Entries

Read more GitHub - slundberg/shap: A unified approach to explain the output of any machine learning model.
0 users, 0 mentions 2018/06/27 10:28
Read more My Internship at Zillow Group AI Part 1: Attribute Recognition in Real Estate Listings - Zillow Tech...
2 users, 1 mentions 2019/09/24 18:20
Read more Explaining Black Box Models: Ensemble and Deep Learning Using LIME and SHAP
0 users, 17 mentions 2020/01/21 15:51
Read more [1911.02508] How can we fool LIME and SHAP? Adversarial Attacks on Post hoc Explanation Methods
0 users, 10 mentions 2019/11/08 02:21