Machine learning feature selection for tuning memory page swapping
MetadataShow full item record
This thesis is an exploration of the virtual memory subsystem in the modern Linux kernel. It applies machine learning to find areas where better page-out decisions can be made. Two areas of possible improvement are identified and analyzed. The first area explored arises because pages in a computation appear repeatedly in a sequence. This is an example of temporal locality. In this instance, we can predict pages that will not be recalled again from the backing store with a precision and recall of 0.82 and 0.81, respectively, with a baseline of 0.30. The second is trying to predict when the system has made bad page-out decisions, those which lived in the backing store for less than one second before being recalled into RAM. In this case, we achieved a precision of 0.82 and a recall of 0.81 with a baseline of 0.12.
Approved for public release; distribution is unlimited
Showing items related by title, author, creator and subject.
Fiebiger, Frank (Monterey, California: Naval Postgraduate School, 2015-09);The digital era has led to an unprecedented increase in the amount of information available, of which an essential part is represented by visual data. The data forensics community asks for machine solutions to face the ...
Nguyen, Doan Han (Monterey, California. Naval Postgraduate School, 1995-12);An important problem in software development process is to make better use of software libraries by improving the search and retrieval process, that is, by making it easier to find the few components you may want among the ...
Pelegris, Gerasimos (Monterey, California. Naval Postgraduate School, 1994-06);ABSTRACT A method for passive ranging based on the principle of triangulation is considered In the basic triangulation scheme that is a single baseline model the precision in the bearing ...