Prediction interval with examples of similar pattern and prediction strength

H. M.Dipu Kabir; Mohammad Anwar Hosen; Saeid Nahavandi; Abbas Khosravi

doi:10.1109/CCECE.2017.7946723

Prediction interval with examples of similar pattern and prediction strength

H. M.Dipu Kabir, Mohammad Anwar Hosen, Saeid Nahavandi, Abbas Khosravi

Research output: Chapter in Book/Conference Proceeding/Report › Conference Paper published in a book › peer-review

13 Citations (Scopus)

Abstract

In this paper, we formed prediction intervals using historical similarities, found through the direct correlation. At first, a string of 5 to 20 recent samples is correlated with a long training string of samples. Then, the highest normalized correlation values and corresponding indexes are picked. After that, the amplitudes of the matched samples are adjusted by multiplying the value with the amplitude of recent string and by dividing by the amplitude of matched strings. These adjusted samples are actually the prediction values. Each prediction value is given a weight (relevance) based on the value of normalized correlation and a function of the ratio between amplitudes of strings. A bar chart is drawn using the weighted (relevance) distribution and less relevant regions are discarded from sides. A prediction strength is calculated from relevances. Except for the calculation of relevance, everything is calculated without any assumption. The user can check similar occurrences and decide to search more when the prediction strength is low.

Original language	English
Title of host publication	2017 IEEE 30th Canadian Conference on Electrical and Computer Engineering, CCECE 2017
Publisher	Institute of Electrical and Electronics Engineers Inc.
ISBN (Electronic)	9781509055388
DOIs	https://doi.org/10.1109/CCECE.2017.7946723
Publication status	Published - 12 Jun 2017
Externally published	Yes
Event	30th IEEE Canadian Conference on Electrical and Computer Engineering, CCECE 2017 - Windsor, Canada Duration: 30 Apr 2017 → 3 May 2017

Publication series

Name	Canadian Conference on Electrical and Computer Engineering
Volume	0
ISSN (Print)	0840-7789

Conference

Conference	30th IEEE Canadian Conference on Electrical and Computer Engineering, CCECE 2017
Country/Territory	Canada
City	Windsor
Period	30/04/17 → 3/05/17

Bibliographical note

Publisher Copyright:
© 2017 IEEE.

Keywords

Correlation
Forecast
Neural Network (NN)
Prediction Interval (PI)
Regression

Access to Document

10.1109/CCECE.2017.7946723

Cite this

Kabir, H. M. D., Hosen, M. A., Nahavandi, S., & Khosravi, A. (2017). Prediction interval with examples of similar pattern and prediction strength. In 2017 IEEE 30th Canadian Conference on Electrical and Computer Engineering, CCECE 2017 Article 7946723 (Canadian Conference on Electrical and Computer Engineering; Vol. 0). Institute of Electrical and Electronics Engineers Inc.. https://doi.org/10.1109/CCECE.2017.7946723

@inproceedings{9197eeb637434726bd54268ff52bdaa2,

title = "Prediction interval with examples of similar pattern and prediction strength",

abstract = "In this paper, we formed prediction intervals using historical similarities, found through the direct correlation. At first, a string of 5 to 20 recent samples is correlated with a long training string of samples. Then, the highest normalized correlation values and corresponding indexes are picked. After that, the amplitudes of the matched samples are adjusted by multiplying the value with the amplitude of recent string and by dividing by the amplitude of matched strings. These adjusted samples are actually the prediction values. Each prediction value is given a weight (relevance) based on the value of normalized correlation and a function of the ratio between amplitudes of strings. A bar chart is drawn using the weighted (relevance) distribution and less relevant regions are discarded from sides. A prediction strength is calculated from relevances. Except for the calculation of relevance, everything is calculated without any assumption. The user can check similar occurrences and decide to search more when the prediction strength is low.",

keywords = "Correlation, Forecast, Neural Network (NN), Prediction Interval (PI), Regression",

author = "Kabir, \{H. M.Dipu\} and Hosen, \{Mohammad Anwar\} and Saeid Nahavandi and Abbas Khosravi",

note = "Publisher Copyright: {\textcopyright} 2017 IEEE.; 30th IEEE Canadian Conference on Electrical and Computer Engineering, CCECE 2017 ; Conference date: 30-04-2017 Through 03-05-2017",

year = "2017",

month = jun,

day = "12",

doi = "10.1109/CCECE.2017.7946723",

language = "English",

series = "Canadian Conference on Electrical and Computer Engineering",

publisher = "Institute of Electrical and Electronics Engineers Inc.",

booktitle = "2017 IEEE 30th Canadian Conference on Electrical and Computer Engineering, CCECE 2017",

}

Kabir, HMD, Hosen, MA, Nahavandi, S & Khosravi, A 2017, Prediction interval with examples of similar pattern and prediction strength. in 2017 IEEE 30th Canadian Conference on Electrical and Computer Engineering, CCECE 2017., 7946723, Canadian Conference on Electrical and Computer Engineering, vol. 0, Institute of Electrical and Electronics Engineers Inc., 30th IEEE Canadian Conference on Electrical and Computer Engineering, CCECE 2017, Windsor, Canada, 30/04/17. https://doi.org/10.1109/CCECE.2017.7946723

Prediction interval with examples of similar pattern and prediction strength. / Kabir, H. M.Dipu; Hosen, Mohammad Anwar; Nahavandi, Saeid et al.
2017 IEEE 30th Canadian Conference on Electrical and Computer Engineering, CCECE 2017. Institute of Electrical and Electronics Engineers Inc., 2017. 7946723 (Canadian Conference on Electrical and Computer Engineering; Vol. 0).

Research output: Chapter in Book/Conference Proceeding/Report › Conference Paper published in a book › peer-review

TY - GEN

T1 - Prediction interval with examples of similar pattern and prediction strength

AU - Kabir, H. M.Dipu

AU - Hosen, Mohammad Anwar

AU - Nahavandi, Saeid

AU - Khosravi, Abbas

PY - 2017/6/12

Y1 - 2017/6/12

N2 - In this paper, we formed prediction intervals using historical similarities, found through the direct correlation. At first, a string of 5 to 20 recent samples is correlated with a long training string of samples. Then, the highest normalized correlation values and corresponding indexes are picked. After that, the amplitudes of the matched samples are adjusted by multiplying the value with the amplitude of recent string and by dividing by the amplitude of matched strings. These adjusted samples are actually the prediction values. Each prediction value is given a weight (relevance) based on the value of normalized correlation and a function of the ratio between amplitudes of strings. A bar chart is drawn using the weighted (relevance) distribution and less relevant regions are discarded from sides. A prediction strength is calculated from relevances. Except for the calculation of relevance, everything is calculated without any assumption. The user can check similar occurrences and decide to search more when the prediction strength is low.

AB - In this paper, we formed prediction intervals using historical similarities, found through the direct correlation. At first, a string of 5 to 20 recent samples is correlated with a long training string of samples. Then, the highest normalized correlation values and corresponding indexes are picked. After that, the amplitudes of the matched samples are adjusted by multiplying the value with the amplitude of recent string and by dividing by the amplitude of matched strings. These adjusted samples are actually the prediction values. Each prediction value is given a weight (relevance) based on the value of normalized correlation and a function of the ratio between amplitudes of strings. A bar chart is drawn using the weighted (relevance) distribution and less relevant regions are discarded from sides. A prediction strength is calculated from relevances. Except for the calculation of relevance, everything is calculated without any assumption. The user can check similar occurrences and decide to search more when the prediction strength is low.

KW - Correlation

KW - Forecast

KW - Neural Network (NN)

KW - Prediction Interval (PI)

KW - Regression

UR - https://www.webofscience.com/wos/woscc/full-record/WOS:000424192700137

UR - https://openalex.org/W2689813587

UR - https://www.scopus.com/pages/publications/85021785217

U2 - 10.1109/CCECE.2017.7946723

DO - 10.1109/CCECE.2017.7946723

M3 - Conference Paper published in a book

AN - SCOPUS:85021785217

T3 - Canadian Conference on Electrical and Computer Engineering

BT - 2017 IEEE 30th Canadian Conference on Electrical and Computer Engineering, CCECE 2017

PB - Institute of Electrical and Electronics Engineers Inc.

T2 - 30th IEEE Canadian Conference on Electrical and Computer Engineering, CCECE 2017

Y2 - 30 April 2017 through 3 May 2017

ER -

Kabir HMD, Hosen MA, Nahavandi S, Khosravi A. Prediction interval with examples of similar pattern and prediction strength. In 2017 IEEE 30th Canadian Conference on Electrical and Computer Engineering, CCECE 2017. Institute of Electrical and Electronics Engineers Inc. 2017. 7946723. (Canadian Conference on Electrical and Computer Engineering). doi: 10.1109/CCECE.2017.7946723

Prediction interval with examples of similar pattern and prediction strength

Abstract

Publication series

Conference

Bibliographical note

Keywords

Access to Document

Other files and links

Fingerprint

Cite this