國立虎尾科技大學 |

Design and Implementation of Inference Services for Random Forests.

Record Type:	Language materials, manuscript : Monograph/item
Title/Author:	Design and Implementation of Inference Services for Random Forests./
Author:	Romero-Gainza, Eduardo A.
Description:	1 online resource (121 pages)
Notes:	Source: Dissertations Abstracts International, Volume: 85-04, Section: B.
Contained By:	Dissertations Abstracts International85-04B.
Subject:	Computer science. -
Online resource:	click for full text (PQDT)
ISBN:	9798380596923

Design and Implementation of Inference Services for Random Forests.
Romero-Gainza, Eduardo A.

Design and Implementation of Inference Services for Random Forests. - 1 online resource (121 pages)

Source: Dissertations Abstracts International, Volume: 85-04, Section: B.

Thesis (Ph.D.)--The Ohio State University, 2023.

Includes bibliographical references

Random forests use ensembles of decision trees to boost accuracy for machine learning tasks. However, large ensembles slow down inferences on platforms that process each tree in an ensemble individually. My initial work has focused on creating Bolt: a platform that restructures random forests, minimizes branch mispredictions and cache misses, and speeds up inference. Bolt is the fastest inference of any model on random forests composed by trees under a certain height. Further, the preponderance of distributed systems and Bolt's novel structures can be leveraged to parallelize forests and achieve even faster inference on random forests. Given a random forest workload and system specifications, we can find the optimal distribution of resources and workload that would lead to the fastest possible inferences. An efficient distributed structure for random forests achieves fast inference on larger forests than previous versions of Bolt. Finally, interpretability is a key advantage of random forests with respect to other machine learning models. We intend to leverage forest interpretability for the field of digital agriculture by providing a mechanism for farmers to evaluate the quality of other models.

Electronic reproduction.
Ann Arbor, Mich. :
ProQuest,
2024

Mode of access: World Wide Web

ISBN: 9798380596923Subjects--Topical Terms:

573171
Computer science.
Subjects--Index Terms:

Inference servicesIndex Terms--Genre/Form:

554714
Electronic books.

Design and Implementation of Inference Services for Random Forests.
LDR:02534ntm a22003737 4500 001 1143892
005 20240517105020.5
006 m o d
007 cr mn ---uuuuu
008 250605s2023 xx obm 000 0 eng d
020 $a 9798380596923
035 $a (MiAaPQ)AAI30788399
035 $a AAI30788399
040 $a MiAaPQ $b eng $c MiAaPQ $d NTU
100 1 $a Romero-Gainza, Eduardo A. $3 1468701
245 1 0 $a Design and Implementation of Inference Services for Random Forests.
264 0 $c 2023
300 $a 1 online resource (121 pages)
336 $a text $b txt $2 rdacontent
337 $a computer $b c $2 rdamedia
338 $a online resource $b cr $2 rdacarrier
500 $a Source: Dissertations Abstracts International, Volume: 85-04, Section: B.
500 $a Advisor: Stewart, Christopher.
502 $a Thesis (Ph.D.)--The Ohio State University, 2023.
504 $a Includes bibliographical references
520 $a Random forests use ensembles of decision trees to boost accuracy for machine learning tasks. However, large ensembles slow down inferences on platforms that process each tree in an ensemble individually. My initial work has focused on creating Bolt: a platform that restructures random forests, minimizes branch mispredictions and cache misses, and speeds up inference. Bolt is the fastest inference of any model on random forests composed by trees under a certain height. Further, the preponderance of distributed systems and Bolt's novel structures can be leveraged to parallelize forests and achieve even faster inference on random forests. Given a random forest workload and system specifications, we can find the optimal distribution of resources and workload that would lead to the fastest possible inferences. An efficient distributed structure for random forests achieves fast inference on larger forests than previous versions of Bolt. Finally, interpretability is a key advantage of random forests with respect to other machine learning models. We intend to leverage forest interpretability for the field of digital agriculture by providing a mechanism for farmers to evaluate the quality of other models.
533 $a Electronic reproduction. $b Ann Arbor, Mich. : $c ProQuest, $d 2024
538 $a Mode of access: World Wide Web
650 4 $a Computer science. $3 573171
650 4 $a Computer engineering. $3 569006
653 $a Inference services
653 $a Digital agriculture
653 $a Random forest workload
655 7 $a Electronic books. $2 local $3 554714
690 $a 0984
690 $a 0464
690 $a 0800
710 2 $a ProQuest Information and Learning Co. $3 1178819
710 2 $a The Ohio State University. $b Computer Science and Engineering. $3 1180873
773 0 $t Dissertations Abstracts International $g 85-04B.
856 4 0 $u http://pqdd.sinica.edu.tw/twdaoapp/servlet/advanced?query=30788399 $z click for full text (PQDT)