Improving Long-Horizon Forecasts with Expectation-Biased LSTM Networks

by   Aya Abdelsalam Ismail, et al.

State-of-the-art forecasting methods using Recurrent Neural Net- works (RNN) based on Long-Short Term Memory (LSTM) cells have shown exceptional performance targeting short-horizon forecasts, e.g given a set of predictor features, forecast a target value for the next few time steps in the future. However, in many applica- tions, the performance of these methods decays as the forecasting horizon extends beyond these few time steps. This paper aims to explore the challenges of long-horizon forecasting using LSTM networks. Here, we illustrate the long-horizon forecasting problem in datasets from neuroscience and energy supply management. We then propose expectation-biasing, an approach motivated by the literature of Dynamic Belief Networks, as a solution to improve long-horizon forecasting using LSTMs. We propose two LSTM ar- chitectures along with two methods for expectation biasing that significantly outperforms standard practice.



There are no comments yet.


page 1

page 2

page 3

page 4


Data-Driven Forecasting of High-Dimensional Chaotic Systems with Long-Short Term Memory Networks

We introduce a data-driven forecasting method for high dimensional, chao...

Forecasting Commodity Prices Using Long Short-Term Memory Neural Networks

This paper applies a recurrent neural network (RNN) method to forecast c...

On Long Memory Origins and Forecast Horizons

Most long memory forecasting studies assume that the memory is generated...

MX-LSTM: mixing tracklets and vislets to jointly forecast trajectories and head poses

Recent approaches on trajectory forecasting use tracklets to predict the...

Conservative Time Discretization: A Comparative Study

We present the first review of methods to overapproximate the set of rea...

Neural networks trained with WiFi traces to predict airport passenger behavior

The use of neural networks to predict airport passenger activity choices...

Multi-time-horizon Solar Forecasting Using Recurrent Neural Network

The non-stationarity characteristic of the solar power renders tradition...
This week in AI

Get the week's most popular data science and artificial intelligence research sent straight to your inbox every Saturday.