UCB (Upper-Confidence-Bound) and Thomson-Sampling analysis for advertisement-selection
Anmol Kumar Soni, A. Boobalan, M. Umapathy, E. Rajesh, Vartika Kesarwani · 2025
Advertising is a marketing communication that engages a disclosed sponsored, open message to promote or sell a certain product, idea, or service. To promote their products or services the sponsors of advertising are often business wishing. For that advertiser pays a big amount of money for advertising their ads on a social network platform. However, the existing strategies use supervised learning which then Using static attribute lists, which become obsolete overtime if not refreshed at regular intervals. That means that when there is a shift in the new data, these techniques are not vulnerable to adaptive learning. Yet most online marketers find it difficult to select the right content that can help achieve their promotional objectives. Using the technique of leveraging learning as it is reward-based links to desired outcomes, that&s;s how the simplification and optimization of the best content for an advertising campaign are done. We will solve a business problem within the digital marketing arena using UCB (Upper-Confidence-Bound) and Thompson-Sampling for selecting the best ads and for identifying the best model between UCB and Thompson-Sampling.