Probabilistic forecasting of symbol sequences with deep neural networks

Michael Hauser, Yiwei Fu, Yue Li, Shashi Phoha, Asok Ray

Research output: Chapter in Book/Report/Conference proceedingConference contribution

1 Citation (Scopus)

Abstract

Time series forecasting is usually done in a deterministic sense, such as in autoregressive moving average models, where a future state is predicted as a linear combination of past events. However, by formulating the problem in a probabilistic sense, soft predictions are obtained from a given probability mass function. This paper uses a deep neural network for probabilistic forecasting of time series by minimizing the cross entropy of the probability of future symbols from a given state. The advantage of this type of model is that it makes probabilistic inferences from the ground up, and without any restrictive assumptions (e.g., second order statistics). The efficacy of the proposed model is tested by forecasting the emergence of combustion instabilities, defined to be the root mean square of the pressure signal inside a laboratory-scale combustor system. The proposed algorithm has been compared with the autoregressive moving average (ARMA) model, which acts as a baseline for many time-series forecasting tasks, and the proposed model is shown to significantly outperform the ARMA model in this task.

Original languageEnglish (US)
Title of host publication2017 American Control Conference, ACC 2017
PublisherInstitute of Electrical and Electronics Engineers Inc.
Pages3147-3152
Number of pages6
ISBN (Electronic)9781509059928
DOIs
StatePublished - Jun 29 2017
Event2017 American Control Conference, ACC 2017 - Seattle, United States
Duration: May 24 2017May 26 2017

Other

Other2017 American Control Conference, ACC 2017
CountryUnited States
CitySeattle
Period5/24/175/26/17

Fingerprint

Time series
Combustors
Deep neural networks
Entropy
Statistics

All Science Journal Classification (ASJC) codes

  • Electrical and Electronic Engineering

Cite this

Hauser, M., Fu, Y., Li, Y., Phoha, S., & Ray, A. (2017). Probabilistic forecasting of symbol sequences with deep neural networks. In 2017 American Control Conference, ACC 2017 (pp. 3147-3152). [7963431] Institute of Electrical and Electronics Engineers Inc.. https://doi.org/10.23919/ACC.2017.7963431
Hauser, Michael ; Fu, Yiwei ; Li, Yue ; Phoha, Shashi ; Ray, Asok. / Probabilistic forecasting of symbol sequences with deep neural networks. 2017 American Control Conference, ACC 2017. Institute of Electrical and Electronics Engineers Inc., 2017. pp. 3147-3152
@inproceedings{3e6efae2d1544c8980f5465bb819a4d4,
title = "Probabilistic forecasting of symbol sequences with deep neural networks",
abstract = "Time series forecasting is usually done in a deterministic sense, such as in autoregressive moving average models, where a future state is predicted as a linear combination of past events. However, by formulating the problem in a probabilistic sense, soft predictions are obtained from a given probability mass function. This paper uses a deep neural network for probabilistic forecasting of time series by minimizing the cross entropy of the probability of future symbols from a given state. The advantage of this type of model is that it makes probabilistic inferences from the ground up, and without any restrictive assumptions (e.g., second order statistics). The efficacy of the proposed model is tested by forecasting the emergence of combustion instabilities, defined to be the root mean square of the pressure signal inside a laboratory-scale combustor system. The proposed algorithm has been compared with the autoregressive moving average (ARMA) model, which acts as a baseline for many time-series forecasting tasks, and the proposed model is shown to significantly outperform the ARMA model in this task.",
author = "Michael Hauser and Yiwei Fu and Yue Li and Shashi Phoha and Asok Ray",
year = "2017",
month = "6",
day = "29",
doi = "10.23919/ACC.2017.7963431",
language = "English (US)",
pages = "3147--3152",
booktitle = "2017 American Control Conference, ACC 2017",
publisher = "Institute of Electrical and Electronics Engineers Inc.",
address = "United States",

}

Hauser, M, Fu, Y, Li, Y, Phoha, S & Ray, A 2017, Probabilistic forecasting of symbol sequences with deep neural networks. in 2017 American Control Conference, ACC 2017., 7963431, Institute of Electrical and Electronics Engineers Inc., pp. 3147-3152, 2017 American Control Conference, ACC 2017, Seattle, United States, 5/24/17. https://doi.org/10.23919/ACC.2017.7963431

Probabilistic forecasting of symbol sequences with deep neural networks. / Hauser, Michael; Fu, Yiwei; Li, Yue; Phoha, Shashi; Ray, Asok.

2017 American Control Conference, ACC 2017. Institute of Electrical and Electronics Engineers Inc., 2017. p. 3147-3152 7963431.

Research output: Chapter in Book/Report/Conference proceedingConference contribution

TY - GEN

T1 - Probabilistic forecasting of symbol sequences with deep neural networks

AU - Hauser, Michael

AU - Fu, Yiwei

AU - Li, Yue

AU - Phoha, Shashi

AU - Ray, Asok

PY - 2017/6/29

Y1 - 2017/6/29

N2 - Time series forecasting is usually done in a deterministic sense, such as in autoregressive moving average models, where a future state is predicted as a linear combination of past events. However, by formulating the problem in a probabilistic sense, soft predictions are obtained from a given probability mass function. This paper uses a deep neural network for probabilistic forecasting of time series by minimizing the cross entropy of the probability of future symbols from a given state. The advantage of this type of model is that it makes probabilistic inferences from the ground up, and without any restrictive assumptions (e.g., second order statistics). The efficacy of the proposed model is tested by forecasting the emergence of combustion instabilities, defined to be the root mean square of the pressure signal inside a laboratory-scale combustor system. The proposed algorithm has been compared with the autoregressive moving average (ARMA) model, which acts as a baseline for many time-series forecasting tasks, and the proposed model is shown to significantly outperform the ARMA model in this task.

AB - Time series forecasting is usually done in a deterministic sense, such as in autoregressive moving average models, where a future state is predicted as a linear combination of past events. However, by formulating the problem in a probabilistic sense, soft predictions are obtained from a given probability mass function. This paper uses a deep neural network for probabilistic forecasting of time series by minimizing the cross entropy of the probability of future symbols from a given state. The advantage of this type of model is that it makes probabilistic inferences from the ground up, and without any restrictive assumptions (e.g., second order statistics). The efficacy of the proposed model is tested by forecasting the emergence of combustion instabilities, defined to be the root mean square of the pressure signal inside a laboratory-scale combustor system. The proposed algorithm has been compared with the autoregressive moving average (ARMA) model, which acts as a baseline for many time-series forecasting tasks, and the proposed model is shown to significantly outperform the ARMA model in this task.

UR - http://www.scopus.com/inward/record.url?scp=85027035902&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=85027035902&partnerID=8YFLogxK

U2 - 10.23919/ACC.2017.7963431

DO - 10.23919/ACC.2017.7963431

M3 - Conference contribution

SP - 3147

EP - 3152

BT - 2017 American Control Conference, ACC 2017

PB - Institute of Electrical and Electronics Engineers Inc.

ER -

Hauser M, Fu Y, Li Y, Phoha S, Ray A. Probabilistic forecasting of symbol sequences with deep neural networks. In 2017 American Control Conference, ACC 2017. Institute of Electrical and Electronics Engineers Inc. 2017. p. 3147-3152. 7963431 https://doi.org/10.23919/ACC.2017.7963431