Aggregated Model Spaces

<a href="https://embed.stanford.edu/iframe/?url=https%3A%2F%2Fpurl.stanford.edu%2Fsd105zh4553" class="su-underline">Show Content</a>

Abstract/Contents

Abstract: In many practical applications, machine learning is divided over multiple agents, where each agent learns a different task and/or learns from a different dataset. We present Aggregated Model Spaces (AMS), a framework for learning a global model by aggregating local learnings performed by each agent. Our approach forgoes sharing of data between agents, makes no assumptions on the distribution of data across agents, and requires minimal communication between agents. We empirically validate our techniques on MNIST experiments and discuss how AMS can generalize to a wide range of problem settings, including federated averaging and catastrophic forgetting. We believe our framework to be among the first to lay out a general methodology for “combining” distinct models.

Type of resource	text
Date created	[ca. May 8, 2018]

Author	Guha, Neel
Degree granting institution	Stanford University, Department of Computer Science
Primary advisor	Sahami, Mehran

Related item	Title Knowledge Aggregation via Epsilon Model Spaces
Location	https://purl.stanford.edu/sd105zh4553

Use and reproduction: User agrees that, where applicable, content will not be used to identify or to otherwise infringe the privacy or confidentiality rights of individuals. Content distributed via the Stanford Digital Repository may be subject to additional license and use restrictions applied by the depositor.
License: This work is licensed under a Creative Commons Attribution Non Commercial 3.0 Unported license (CC BY-NC).

Preferred Citation: Guha, Neel. (2018). Aggregated Model Spaces. Stanford Digital Repository. Available at: https://purl.stanford.edu/sd105zh4553

Undergraduate Theses, School of Engineering

Loading usage metrics...