Modeling Long- and Short-Term Temporal Patterns with Deep Neural Networks (2017-03-21T00:00:00.000000Z)

TL;DR

A novel deep learning framework, namely Long- and Short-term Time-series network (LSTNet), to address this open challenge of multivariate time series forecasting, using the Convolution Neural Network and the Recurrent Neural Network to extract short-term local dependency patterns among variables and to discover long-term patterns for time series trends.

Abstract

Multivariate time series forecasting is an important machine learning problem across many domains, including predictions of solar plant energy output, electricity consumption, and traffic jam situation. Temporal data arise in these real-world applications often involves a mixture of long-term and short-term patterns, for which traditional approaches such as Autoregressive models and Gaussian Process may fail. In this paper, we proposed a novel deep learning framework, namely Long- and Short-term Time-series network (LSTNet), to address this open challenge. LSTNet uses the Convolution Neural Network (CNN) and the Recurrent Neural Network (RNN) to extract short-term local dependency patterns among variables and to discover long-term patterns for time series trends. Furthermore, we leverage traditional autoregressive model to tackle the scale insensitive problem of the neural network model. In our evaluation on real-world data with complex mixtures of repetitive patterns, LSTNet achieved significant performance improvements over that of several state-of-the-art baseline methods. All the data and experiment codes are available online.

Authors

Wei-Cheng Chang

6 papers

Yiming Yang

11 papers

Guokun Lai

3 papers

TL;DR

Abstract

Authors

References38 items

What to Do Next: Modeling User Behaviors by Time-LSTM

Nonlinear Dynamic Boltzmann Machines for Time-Series Prediction

Temporal Regularized Matrix Factorization for High-dimensional Time Series Prediction

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

Temporal Convolutional Networks: A Unified Approach to Action Segmentation

Recurrent Neural Networks for Multivariate Time Series with Missing Values

Deep, Convolutional, and Recurrent Models for Human Activity Recognition Using Wearables

Learning to Diagnose with LSTM Recurrent Neural Networks

Bayesian Time Series Learning with Gaussian Processes

Deep Convolutional Neural Networks on Multichannel Time Series for Human Activity Recognition

Robust Estimation of Transition Matrices in High Dimensional Heavy-tailed Vector Autoregressive Processes

Highway Networks

Adam: A Method for Stochastic Optimization

Empirical Evaluation of Gated Recurrent Neural Networks on Sequence Modeling

Neural Machine Translation by Jointly Learning to Align and Translate

Forecasting Macroeconomic Time Series: LASSO-Based Approaches and Their Forecast Combinations with Dynamic Factor Models

Bayesian Inference and Learning in Gaussian Process State-Space Models with Particle MCMC

Gaussian processes for time-series modelling

ImageNet classification with deep convolutional neural networks

Et al

New Introduction to Multiple Time Series Analysis

Hybrid neural network models for hydrologic time series forecasting

Support vector machine with adaptive parameters in financial time series forecasting

Financial time series forecasting using support vector machines

Convolutional networks for images, speech, and time series

FORECASTING WITH ARTIFICIAL NEURAL NETWORKS: THE STATE OF THE ART

Long Short-Term Memory

Support Vector Method for Function Approximation, Regression Estimation and Signal Processing

Recurrent Networks and NARMA Modeling

Finding Structure in Time

General exponential smoothing and the equivalent arma process

Time Series Analysis: Forecasting and Control

Distribution of Residual Autocorrelations in Autoregressive-Integrated Moving Average Time Series Models

Deep Learning: A Generic Approach for Extreme Condition Traffic Forecasting

FUNDAMENTAL TECHNOLOGIES IN MODERN SPEECH RECOGNITION Digital Object Identifier 10.1109/MSP.2012.2205597

Time series forecasting using a hybrid ARIMA and neural network model

The State of the Art

Supplement: Estimating Structured Vector Autoregressive Model

Field of Study

Journal Information

Name

Venue Information

Name

Type

URL

Alternate Names