Causal inference via sparse additive models with application to online advertising

Wei Sun, Pengyuan Wang, Dawei Yin, Jian Yang, Yi Chang

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

9 Citations (Scopus)

Abstract

Advertising effectiveness measurement is a fundamental problem in online advertising. Various causal inference methods have been employed to measure the causal effects of ad treatments. However, existing methods mainly focus on linear logistic regression for univariate and binary treatments and are not well suited for complex ad treatments of multi-dimensions, where each dimension could be discrete or continuous. In this paper we propose a novel two-stage causal inference framework for assessing the impact of complex ad treatments. In the first stage, we estimate the propensity parameter via a sparse additive model; in the second stage, a propensity-adjusted regression model is applied for measuring the treatment effect. Our approach is shown to provide an unbiased estimation of the ad effectiveness under regularity conditions. To demonstrate the efficacy of our approach, we apply it to a real online advertising campaign to evaluate the impact of three ad treatments: ad frequency, ad channel, and ad size. We show that the ad frequency usually has a treatment effect cap when ads are showing on mobile device. In addition, the strategies for choosing best ad size are completely different for mobile ads and online ads.

Original languageEnglish (US)
Title of host publicationProceedings of the 29th AAAI Conference on Artificial Intelligence, AAAI 2015 and the 27th Innovative Applications of Artificial Intelligence Conference, IAAI 2015
PublisherAI Access Foundation
Pages297-303
Number of pages7
Volume1
ISBN (Electronic)9781577356998
StatePublished - Jun 1 2015
Externally publishedYes
Event29th AAAI Conference on Artificial Intelligence, AAAI 2015 and the 27th Innovative Applications of Artificial Intelligence Conference, IAAI 2015 - Austin, United States
Duration: Jan 25 2015Jan 30 2015

Other

Other29th AAAI Conference on Artificial Intelligence, AAAI 2015 and the 27th Innovative Applications of Artificial Intelligence Conference, IAAI 2015
CountryUnited States
CityAustin
Period1/25/151/30/15

Fingerprint

Marketing
Mobile devices
Logistics

ASJC Scopus subject areas

  • Software
  • Artificial Intelligence

Cite this

Sun, W., Wang, P., Yin, D., Yang, J., & Chang, Y. (2015). Causal inference via sparse additive models with application to online advertising. In Proceedings of the 29th AAAI Conference on Artificial Intelligence, AAAI 2015 and the 27th Innovative Applications of Artificial Intelligence Conference, IAAI 2015 (Vol. 1, pp. 297-303). AI Access Foundation.

Causal inference via sparse additive models with application to online advertising. / Sun, Wei; Wang, Pengyuan; Yin, Dawei; Yang, Jian; Chang, Yi.

Proceedings of the 29th AAAI Conference on Artificial Intelligence, AAAI 2015 and the 27th Innovative Applications of Artificial Intelligence Conference, IAAI 2015. Vol. 1 AI Access Foundation, 2015. p. 297-303.

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

Sun, W, Wang, P, Yin, D, Yang, J & Chang, Y 2015, Causal inference via sparse additive models with application to online advertising. in Proceedings of the 29th AAAI Conference on Artificial Intelligence, AAAI 2015 and the 27th Innovative Applications of Artificial Intelligence Conference, IAAI 2015. vol. 1, AI Access Foundation, pp. 297-303, 29th AAAI Conference on Artificial Intelligence, AAAI 2015 and the 27th Innovative Applications of Artificial Intelligence Conference, IAAI 2015, Austin, United States, 1/25/15.
Sun W, Wang P, Yin D, Yang J, Chang Y. Causal inference via sparse additive models with application to online advertising. In Proceedings of the 29th AAAI Conference on Artificial Intelligence, AAAI 2015 and the 27th Innovative Applications of Artificial Intelligence Conference, IAAI 2015. Vol. 1. AI Access Foundation. 2015. p. 297-303
Sun, Wei ; Wang, Pengyuan ; Yin, Dawei ; Yang, Jian ; Chang, Yi. / Causal inference via sparse additive models with application to online advertising. Proceedings of the 29th AAAI Conference on Artificial Intelligence, AAAI 2015 and the 27th Innovative Applications of Artificial Intelligence Conference, IAAI 2015. Vol. 1 AI Access Foundation, 2015. pp. 297-303
@inproceedings{bb77e8f14cd743faa0ef44c1036cbafd,
title = "Causal inference via sparse additive models with application to online advertising",
abstract = "Advertising effectiveness measurement is a fundamental problem in online advertising. Various causal inference methods have been employed to measure the causal effects of ad treatments. However, existing methods mainly focus on linear logistic regression for univariate and binary treatments and are not well suited for complex ad treatments of multi-dimensions, where each dimension could be discrete or continuous. In this paper we propose a novel two-stage causal inference framework for assessing the impact of complex ad treatments. In the first stage, we estimate the propensity parameter via a sparse additive model; in the second stage, a propensity-adjusted regression model is applied for measuring the treatment effect. Our approach is shown to provide an unbiased estimation of the ad effectiveness under regularity conditions. To demonstrate the efficacy of our approach, we apply it to a real online advertising campaign to evaluate the impact of three ad treatments: ad frequency, ad channel, and ad size. We show that the ad frequency usually has a treatment effect cap when ads are showing on mobile device. In addition, the strategies for choosing best ad size are completely different for mobile ads and online ads.",
author = "Wei Sun and Pengyuan Wang and Dawei Yin and Jian Yang and Yi Chang",
year = "2015",
month = "6",
day = "1",
language = "English (US)",
volume = "1",
pages = "297--303",
booktitle = "Proceedings of the 29th AAAI Conference on Artificial Intelligence, AAAI 2015 and the 27th Innovative Applications of Artificial Intelligence Conference, IAAI 2015",
publisher = "AI Access Foundation",

}

TY - GEN

T1 - Causal inference via sparse additive models with application to online advertising

AU - Sun, Wei

AU - Wang, Pengyuan

AU - Yin, Dawei

AU - Yang, Jian

AU - Chang, Yi

PY - 2015/6/1

Y1 - 2015/6/1

N2 - Advertising effectiveness measurement is a fundamental problem in online advertising. Various causal inference methods have been employed to measure the causal effects of ad treatments. However, existing methods mainly focus on linear logistic regression for univariate and binary treatments and are not well suited for complex ad treatments of multi-dimensions, where each dimension could be discrete or continuous. In this paper we propose a novel two-stage causal inference framework for assessing the impact of complex ad treatments. In the first stage, we estimate the propensity parameter via a sparse additive model; in the second stage, a propensity-adjusted regression model is applied for measuring the treatment effect. Our approach is shown to provide an unbiased estimation of the ad effectiveness under regularity conditions. To demonstrate the efficacy of our approach, we apply it to a real online advertising campaign to evaluate the impact of three ad treatments: ad frequency, ad channel, and ad size. We show that the ad frequency usually has a treatment effect cap when ads are showing on mobile device. In addition, the strategies for choosing best ad size are completely different for mobile ads and online ads.

AB - Advertising effectiveness measurement is a fundamental problem in online advertising. Various causal inference methods have been employed to measure the causal effects of ad treatments. However, existing methods mainly focus on linear logistic regression for univariate and binary treatments and are not well suited for complex ad treatments of multi-dimensions, where each dimension could be discrete or continuous. In this paper we propose a novel two-stage causal inference framework for assessing the impact of complex ad treatments. In the first stage, we estimate the propensity parameter via a sparse additive model; in the second stage, a propensity-adjusted regression model is applied for measuring the treatment effect. Our approach is shown to provide an unbiased estimation of the ad effectiveness under regularity conditions. To demonstrate the efficacy of our approach, we apply it to a real online advertising campaign to evaluate the impact of three ad treatments: ad frequency, ad channel, and ad size. We show that the ad frequency usually has a treatment effect cap when ads are showing on mobile device. In addition, the strategies for choosing best ad size are completely different for mobile ads and online ads.

UR - http://www.scopus.com/inward/record.url?scp=84959516996&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84959516996&partnerID=8YFLogxK

M3 - Conference contribution

VL - 1

SP - 297

EP - 303

BT - Proceedings of the 29th AAAI Conference on Artificial Intelligence, AAAI 2015 and the 27th Innovative Applications of Artificial Intelligence Conference, IAAI 2015

PB - AI Access Foundation

ER -