An Investigation on Online Versus Batch Learning in Predicting User Behaviour

Nikolay Burlutskiy, Miltiadis Petridis, Andrew Fish, Alexey Chernov, Nour Ali

Research output: Chapter in Book/Conference proceeding with ISSN or ISBNConference contribution with ISSN or ISBNResearchpeer-review

Abstract

An investigation on how to produce a fast and accurate prediction of user behaviour on the Web is conducted. First, the problem of predicting user behaviour as a classification task is formulated and then the main problems of such real-time predictions are specified: the accuracy and time complexity of the prediction. Second, a method for comparison of online and batch (offline) algorithms used for user behaviour prediction is proposed. Last, the performance of these algorithms using the data from a popular question and answer platform, Stack Overflow, is empirically explored. It is demonstrated that a simple online learning algorithm outperforms state-of-the-art batch algorithms and performs as well as a deep learning algorithm, Deep Belief Networks. The proposed method for comparison of online and offline algorithms as well as the provided experimental evidence can be used for choosing a machine learning set-up for predicting user behaviour on the Web in scenarios where the accuracy and the time performance are of main concern.
Original languageEnglish
Title of host publicationInternational Conference on Innovative Techniques and Applications of Artificial Intelligence
Place of PublicationCambridge, UK
PublisherSpringer International Publishing
Pages135-149
Number of pages15
ISBN (Electronic)9783319471754
ISBN (Print)9783319471747
DOIs
Publication statusPublished - 5 Nov 2016
EventInternational Conference on Innovative Techniques and Applications of Artificial Intelligence - Cambridge, 13-15 December 2016
Duration: 5 Nov 2016 → …

Conference

ConferenceInternational Conference on Innovative Techniques and Applications of Artificial Intelligence
Period5/11/16 → …

Fingerprint

Learning algorithms
Bayesian networks
Learning systems
Deep learning

Bibliographical note

The final publication is available at Springer via http://dx.doi.org/10.1007/978-3-319-47175-4_9

Keywords

  • Online Learning
  • Deep Learning
  • Classification

Cite this

Burlutskiy, N., Petridis, M., Fish, A., Chernov, A., & Ali, N. (2016). An Investigation on Online Versus Batch Learning in Predicting User Behaviour. In International Conference on Innovative Techniques and Applications of Artificial Intelligence (pp. 135-149). Cambridge, UK: Springer International Publishing. https://doi.org/10.1007/978-3-319-47175-4_9
Burlutskiy, Nikolay ; Petridis, Miltiadis ; Fish, Andrew ; Chernov, Alexey ; Ali, Nour. / An Investigation on Online Versus Batch Learning in Predicting User Behaviour. International Conference on Innovative Techniques and Applications of Artificial Intelligence. Cambridge, UK : Springer International Publishing, 2016. pp. 135-149
@inproceedings{5485620e14fe4101858b72d8adc5185f,
title = "An Investigation on Online Versus Batch Learning in Predicting User Behaviour",
abstract = "An investigation on how to produce a fast and accurate prediction of user behaviour on the Web is conducted. First, the problem of predicting user behaviour as a classification task is formulated and then the main problems of such real-time predictions are specified: the accuracy and time complexity of the prediction. Second, a method for comparison of online and batch (offline) algorithms used for user behaviour prediction is proposed. Last, the performance of these algorithms using the data from a popular question and answer platform, Stack Overflow, is empirically explored. It is demonstrated that a simple online learning algorithm outperforms state-of-the-art batch algorithms and performs as well as a deep learning algorithm, Deep Belief Networks. The proposed method for comparison of online and offline algorithms as well as the provided experimental evidence can be used for choosing a machine learning set-up for predicting user behaviour on the Web in scenarios where the accuracy and the time performance are of main concern.",
keywords = "Online Learning, Deep Learning, Classification",
author = "Nikolay Burlutskiy and Miltiadis Petridis and Andrew Fish and Alexey Chernov and Nour Ali",
note = "The final publication is available at Springer via http://dx.doi.org/10.1007/978-3-319-47175-4_9",
year = "2016",
month = "11",
day = "5",
doi = "10.1007/978-3-319-47175-4_9",
language = "English",
isbn = "9783319471747",
pages = "135--149",
booktitle = "International Conference on Innovative Techniques and Applications of Artificial Intelligence",
publisher = "Springer International Publishing",

}

Burlutskiy, N, Petridis, M, Fish, A, Chernov, A & Ali, N 2016, An Investigation on Online Versus Batch Learning in Predicting User Behaviour. in International Conference on Innovative Techniques and Applications of Artificial Intelligence. Springer International Publishing, Cambridge, UK, pp. 135-149, International Conference on Innovative Techniques and Applications of Artificial Intelligence, 5/11/16. https://doi.org/10.1007/978-3-319-47175-4_9

An Investigation on Online Versus Batch Learning in Predicting User Behaviour. / Burlutskiy, Nikolay; Petridis, Miltiadis; Fish, Andrew; Chernov, Alexey; Ali, Nour.

International Conference on Innovative Techniques and Applications of Artificial Intelligence. Cambridge, UK : Springer International Publishing, 2016. p. 135-149.

Research output: Chapter in Book/Conference proceeding with ISSN or ISBNConference contribution with ISSN or ISBNResearchpeer-review

TY - GEN

T1 - An Investigation on Online Versus Batch Learning in Predicting User Behaviour

AU - Burlutskiy, Nikolay

AU - Petridis, Miltiadis

AU - Fish, Andrew

AU - Chernov, Alexey

AU - Ali, Nour

N1 - The final publication is available at Springer via http://dx.doi.org/10.1007/978-3-319-47175-4_9

PY - 2016/11/5

Y1 - 2016/11/5

N2 - An investigation on how to produce a fast and accurate prediction of user behaviour on the Web is conducted. First, the problem of predicting user behaviour as a classification task is formulated and then the main problems of such real-time predictions are specified: the accuracy and time complexity of the prediction. Second, a method for comparison of online and batch (offline) algorithms used for user behaviour prediction is proposed. Last, the performance of these algorithms using the data from a popular question and answer platform, Stack Overflow, is empirically explored. It is demonstrated that a simple online learning algorithm outperforms state-of-the-art batch algorithms and performs as well as a deep learning algorithm, Deep Belief Networks. The proposed method for comparison of online and offline algorithms as well as the provided experimental evidence can be used for choosing a machine learning set-up for predicting user behaviour on the Web in scenarios where the accuracy and the time performance are of main concern.

AB - An investigation on how to produce a fast and accurate prediction of user behaviour on the Web is conducted. First, the problem of predicting user behaviour as a classification task is formulated and then the main problems of such real-time predictions are specified: the accuracy and time complexity of the prediction. Second, a method for comparison of online and batch (offline) algorithms used for user behaviour prediction is proposed. Last, the performance of these algorithms using the data from a popular question and answer platform, Stack Overflow, is empirically explored. It is demonstrated that a simple online learning algorithm outperforms state-of-the-art batch algorithms and performs as well as a deep learning algorithm, Deep Belief Networks. The proposed method for comparison of online and offline algorithms as well as the provided experimental evidence can be used for choosing a machine learning set-up for predicting user behaviour on the Web in scenarios where the accuracy and the time performance are of main concern.

KW - Online Learning

KW - Deep Learning

KW - Classification

U2 - 10.1007/978-3-319-47175-4_9

DO - 10.1007/978-3-319-47175-4_9

M3 - Conference contribution with ISSN or ISBN

SN - 9783319471747

SP - 135

EP - 149

BT - International Conference on Innovative Techniques and Applications of Artificial Intelligence

PB - Springer International Publishing

CY - Cambridge, UK

ER -

Burlutskiy N, Petridis M, Fish A, Chernov A, Ali N. An Investigation on Online Versus Batch Learning in Predicting User Behaviour. In International Conference on Innovative Techniques and Applications of Artificial Intelligence. Cambridge, UK: Springer International Publishing. 2016. p. 135-149 https://doi.org/10.1007/978-3-319-47175-4_9