Using boosted trees for click-through rate prediction for sponsored search

Trofimov, Ilya; Kornetova, A. N.; Topinskiy, V. A.

doi:10.1145/2351356.2351358

Cited by 41 publications

(25 citation statements)

References 5 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Vowpal Wabbit is particularly suitable for large scale learning, since it supports a variety of infrastructures, from single CPUs to compute clusters. As nonlinear regressors we train gradient boosted regression trees using the MatrixNet package [31] with default parameters (500 oblivious trees, depth 6, learning rate 0.1).…”

Section: Methodsmentioning

confidence: 99%

Closed-Form Approximate CRF Training for Scalable Image Segmentation

Kolesnikov

Guillaumin

Ferrari

et al. 2014

Computer Vision – ECCV 2014

View full text Add to dashboard Cite

Abstract. We present LS-CRF, a new method for training cyclic Conditional Random Fields (CRFs) from large datasets that is inspired by classical closed-form expressions for the maximum likelihood parameters of a generative graphical model with tree topology. Training a CRF with LS-CRF requires only solving a set of independent regression problems, each of which can be solved efficiently in closed form or by an iterative solver. This makes LS-CRF orders of magnitude faster than classical CRF training based on probabilistic inference, and at the same time more flexible and easier to implement than other approximate techniques, such as pseudolikelihood or piecewise training. We apply LS-CRF to the task of semantic image segmentation, showing that it achieves on par accuracy to other training techniques at higher speed, thereby allowing efficient CRF training from very large training sets. For example, training a linearly parameterized pairwise CRF on 150,000 images requires less than one hour on a modern workstation.

show abstract

Section: Methodsmentioning

confidence: 99%

Closed-Form Approximate CRF Training for Scalable Image Segmentation

Kolesnikov

Guillaumin

Ferrari

et al. 2014

Computer Vision – ECCV 2014

View full text Add to dashboard Cite

show abstract

“…By contrast, non-linear models are capable of learning feature interactions in various ways and could potentially improve prediction performance [24,31]. Gradient boosting decision trees (GBDT) [31,15] are a straightforward non-linear model to capture feature interactions. Moreover, latent factor models, particularly factorization machines (FMs) [24], map each binary feature into a low dimensional continuous space, and the feature interaction is automatically explored via vector inner product.…”

Section: Related Workmentioning

confidence: 99%

“…Bidding strategies are normally built by estimating the utility of each ad impression, which is commonly done by predicting the underlying user's click-through rate (CTR) or conversion rate (CVR) [17]. Existing solutions for predicting CTR range from linear models [13,21,15,17,26] and gradient boosting decision trees (GBDT) [31] to factorization machines [24]. All of them aim to return a predicted CTR value of the given ad impression.…”

Section: Introductionmentioning

confidence: 99%

Managing Risk of Bidding in Display Advertising

Zhang

Rong

et al. 2017

Proceedings of the Tenth ACM International Conference on Web Search and Data Mining

View full text Add to dashboard Cite

In this paper, we deal with the uncertainty of bidding for display advertising. Similar to the financial market trading, real-time bidding (RTB) based display advertising employs an auction mechanism to automate the impression level media buying; and running a campaign is no different than an investment of acquiring new customers in return for obtaining additional converted sales. Thus, how to optimally bid on an ad impression to drive the profit and return-on-investment becomes essential. However, the large randomness of the user behaviors and the cost uncertainty caused by the auction competition may result in a significant risk from the campaign performance estimation. In this paper, we explicitly model the uncertainty of user click-through rate estimation and auction competition to capture the risk. We borrow an idea from finance and derive the value at risk for each ad display opportunity. Our formulation results in two risk-aware bidding strategies that penalize risky ad impressions and focus more on the ones with higher expected return and lower risk. The empirical study on real-world data demonstrates the effectiveness of our proposed risk-aware bidding strategies: yielding profit gains of 15.4% in offline experiments and up to 17.5% in an online A/B test on a commercial RTB platform over the widely applied bidding strategies

show abstract

“…Achieving the highest click prediction rate means that these agencies can pay to place their online advertisements effectively to target people most interested in their product. Most existing approaches are based on logistic regression or regression tree models (Trofi mov, Kornetova, & Topinskiy, 2012). The model based on deep learning will be discussed to predict the click rate.…”

Section: Introductionmentioning

confidence: 99%

Predicting Advertisement Clicks Using Deep Networks: Interpreting Deep Learning Models

Samel¹

2017

JPUR

View full text Add to dashboard Cite

Using boosted trees for click-through rate prediction for sponsored search

Cited by 41 publications

References 5 publications

Closed-Form Approximate CRF Training for Scalable Image Segmentation

Closed-Form Approximate CRF Training for Scalable Image Segmentation

Managing Risk of Bidding in Display Advertising

Predicting Advertisement Clicks Using Deep Networks: Interpreting Deep Learning Models

Contact Info

Product

Resources

About