Temporal Fusion Transformers for Interpretable Multi-horizon Time Series Forecasting (2019-12-19T00:00:00.000000Z)

TL;DR

The Temporal Fusion Transformer is introduced -- a novel attention-based architecture which combines high-performance multi-horizon forecasting with interpretable insights into temporal dynamics and three practical interpretability use-cases of TFT are showcased.

Authors

Bryan Lim

3 papers

Sercan Ö. Arik

9 papers

Tomas Pfister

7 papers

Nicolas Loeff

1 papers

TL;DR

Authors

References44 items

Documenting evidence of a reuse of ‘“why should I trust you?”: explaining the predictions of any classifier’

RL-LIM: Reinforcement Learning-based Locally Interpretable Modeling

TabNet: Attentive Interpretable Tabular Learning

Multi-Horizon Time Series Forecasting with Temporal Attention Learning

Enhancing the Locality and Breaking the Memory Bottleneck of Transformer on Time Series Forecasting

Deep Factors for Forecasting

Exploring Interpretable LSTM Neural Networks over Multi-Variable Data

Multi-step prediction for influenza outbreak by an adjusted long short-term memory

A Multi-Horizon Quantile Recurrent Forecaster

Attend and Diagnose: Clinical Time Series Analysis using Attention Models

Probabilistic Demand Forecasting at Scale

Attention is All you Need

A Unified Approach to Interpreting Model Predictions

Residual Attention Network for Image Classification

DeepAR: Probabilistic Forecasting with Autoregressive Recurrent Networks

Language Modeling with Gated Convolutional Networks

Temporal Regularized Matrix Factorization for High-dimensional Time Series Prediction

Phased LSTM: Accelerating Recurrent Network Training for Long or Event-based Sequences

RETAIN: An Interpretable Predictive Model for Healthcare using Reverse Time Attention Mechanism

Layer Normalization

“Why Should I Trust You?”: Explaining the Predictions of Any Classifier

A Theoretically Grounded Application of Dropout in Recurrent Neural Networks

Fast and Accurate Deep Network Learning by Exponential Linear Units (ELUs)

The Turn-of-the-Month-Effect: Evidence from Periodic Generalized Autoregressive Conditional Heteroskedasticity (PGARCH) Model

A Genotype-First Approach to Defining the Subtypes of a Complex Disease

A Clockwork RNN

Distributed Lags and Dynamic Models

Regime Changes and Financial Markets

Multiple-output modeling for multi-step-ahead time series forecasting

Multi-horizon inflation forecasts using disaggregated data

A Comparison of Direct and Iterated Multistep Ar Methods for Forecasting Macroeconomic Time Series

Kernel-Based Object Tracking

Timing of Seasonal Sales

Long Short-Term Memory

The Divergence and Bhattacharyya Distance Measures in Signal Selection

The M4 Competition: 100,000 time series and 61 forecasting methods

Attentive State-Space Modeling of Disease Progression

Deep State Space Models for Time Series Forecasting

Temporal Causal Inference with Time Lag

Forecasting Treatment Responses Over Time Using Recurrent Marginal Structural Networks

CorporacionFavoritaGrocerySalesForecastingCom-petition

Oxford- Man Institute’s Realized Library

Modelling Seasonality

and scenarios. Gated Linear Units extensively are utilized throughout our architecture, and Gated Residual Network is proposed as a main building block

Field of Study

Journal Information

Name

Volume

Venue Information

Name

Type

URL

Alternate Names