Programmatic Marketing via Reinforcement Learning
Journal Title: Journal of Business Management & Economics - Year 2017, Vol 5, Issue 11
Abstract
This new algorithm which incorporates look-ahead search inside the training loop resulting in rapid improvement and precise and stable learning. By using this new search methods towards programmatic marketing, we massively improve the return of investment on Google Adwords auction by 4.6% in 1 day. We apply simple, gradient-based updates to train the next policy and value network. This appears to be much more stable than incremental, gradient-based policy improvements such as policy gradient or Q-learning that can potentially forget previous improvements.
Authors and Affiliations
Mengmeng Chen, Luis Rabelo
Keywords
Related Articles
- EP ID EP321302
- DOI 10.15520/jbme.2017.vol5.iss11.269.pp01-04
- Views 63
- Downloads 0
Housing Price Leads Rent in Hong Kong? -- A Puzzle of Rent-Price Theory
This study analyzes the lead-lag relationships between housing price growths, rent growths and the expected housing rent changes. Theories, including the Di Pasquale-Wheaton (DW) Model, assumes housing price as a discoun...
Impact of Capital Structure on Insurance Firms’ Perfromance in Nigeria
The study investigates the impact of capital structure on firm’s performance (Insurance companies) in Nigeria for a period of 5 years (2011-2015) for a sample size of 12 insurance companies. Data were collected from the...
The Impact of Transatlantic Trade and Investment Partnership (Ttip) On the usa and European Union Strategic Cooperation
The main aim of the paper is presentation of the impact of TTIP on the USA and European Union strategic cooperation. It has been emphasized that the talks will make reducing regulatory barriers. New agreements to remove...
Export Competitiveness of Sugarcane in Southern India
This paper combines policy analysis matrix techniques to model the analysis of profitability from sugarcane cultivation. Policy analysis matrices are computed for a sample of sugarcane growers located in the wet land of...
Optimum Executive Body Has 5 Members Simple Mathematical Proof
Mathematical proof: Optimum executive body has 5 members.