Off-campus UMass Amherst users: To download campus access dissertations, please use the following link to log into our proxy server with your UMass Amherst user name and password.

Non-UMass Amherst users: Please talk to your librarian about requesting this dissertation through interlibrary loan.

Dissertations that have an embargo placed on them will not be available to anyone until the embargo expires.

Author ORCID Identifier


Open Access Dissertation

Document Type


Degree Name

Doctor of Philosophy (PhD)

Degree Program

Computer Science

Year Degree Awarded


Month Degree Awarded


First Advisor

J. Eliot B. Moss

Second Advisor

Benjamin Marlin

Third Advisor

Philip Thomas

Fourth Advisor

Antony Hosking

Subject Categories

Artificial Intelligence and Robotics


Garbage collectors are nearly ubiquitous in modern programming languages, and we want to minimize the cost they impose in terms of time and space. Generally, a collector waits until its space is full and then performs a collection to reclaim needed memory. However, this is not the only option; a collection could be performed early when some free space remains. For copying collectors, which are what we consider here, the system must traverse the graph of live objects and copy them, so the cost of a collection is proportional to the volume of objects that are live. Since this value fluctuates during a program's execution, a collector can minimize its cost by carefully choosing the points at which it collects. We help to realize this goal in two ways. First, by developing an algorithm that analyzes after-the-fact traces and computes optimal collection points, we can explore the theoretical limits of garbage collectors. This gives insight into what performance gains are possible, and can guide future collector development into areas that could be most fruitful. Second, we use techniques from machine learning to find improved garbage collection policies that could be implemented in real systems. The optimal collection points provide ground truth from which a model can learn.


Creative Commons License

Creative Commons Attribution-Noncommercial 4.0 License
This work is licensed under a Creative Commons Attribution-Noncommercial-Share Alike 4.0 License.