Accession Number:

AD1111900

Title:

Topics in Advanced Computing: Promise and Challenges of Recommendation Systems for the DoD

Descriptive Note:

Technical Report

Corporate Author:

CARNEGIE-MELLON UNIV PITTSBURGH PA PITTSBURGH United States

Report Date:

2020-01-01

Pagination or Media Count:

21.0

Abstract:

What is a Recommendation System Given your profile and the things youve liked in the past, what is the probability that you will click through on a recommendation Netflix, Amazon, YouTube, Spotify, Facebook, Twitter. DNN-based personalized recommendation models comprise up to 79 percent of AI inference cycles in a production-scale data center. The Idea Behind Recommendation Systems. Given a user and an item that the user has not interacted with, what is the probability that the user will click on the item User-item pairs with the highest predicted click-through rate are prioritized. The data is sparse, i.e., any given user has interacted with very few items.

Subject Categories:

  • Computer Programming and Software
  • Computer Systems Management and Standards

Distribution Statement:

APPROVED FOR PUBLIC RELEASE