Amazon Technologies, Inc.

Yi Liu

Vishwanathan Swaminathan

Jian Liu

Sriram Srinivasan

Houssam Nassif

Tao Hu

Anand Mahadeva Iyer

Rahul Makhijani

Wenjun Zeng

Dale Struble

Hau Wing Calvin Kwok

Shreya Chakrabarti

Daniel Hill

Zachary Wake Austin

Technologies are provided for the generating of optimal policies for bidding in auctions having unknown dynamics. In some embodiments, a computing system can configure many multi-armed bandit (MAB) models defining candidate directed contents for a sequence of pages. A particular MAB model of the many MAB models defines candidate directed contents for a particular page in the sequence of pages, where each arm in the particular MAB model corresponds to a candidate impression on the particular page. The computing system can then determine a solution to an optimization problem with respect to an objective function based on an expected long-term reward for a defined impression on the first page, a defined impression on the second page, and a defined impression on the third page. The solution results in respective directed content for presentation on the first, second, and third pages.

Directed content recommendations for a sequence of multiple pages

Techniques are provided for optimizing a graph. The graph can be generated with a source node, a plurality of item nodes, and a terminal node. Once generated, the graph can be optimized to determine a lowest cost flow from the source node to the terminal node based at least in part on respective budget measurements. A first account node and a first item node can be determined and an account corresponding to the first account node, as well as information that identifies an item corresponding to the first item node can be transmitted.

Method, system, and manufacture for min-cost flow item recommendations

An optimal combination of content items may be determined and served to increase the likelihood of a predefined user interaction. A user can request content which may include an entity (such as a web page, document, advertisement, and the like) that has multiple components through which content items may be provided. For example, a web page may include multiple slots where content items may be displayed on the web page. Each component may be associated with multiple possible content items, resulting in many combinations of layouts for an entity. A content server may determine which layout to provide using a content selection model that is weighted based on a likelihood of groupings of content items resulting in a user interaction that satisfies a success condition (e.g., selecting a hyperlink, selecting a content item, initiating a transaction, etc.).

Artificial intelligence system for optimizing network-accessible content

Growing community of inventors

Seattle, WA, United States of America

Yi Liu