公开(公告)号：US10216899B2

公开(公告)日：2019-02-26

申请号：US15298412

申请日：2016-10-20

Applicant: Hewlett Packard Enterprise Development LP

Inventor： Mehran Kafai ,  Kari Lam

IPC: G06F17/30 ,  G06F19/28 ,  G06N99/00

Abstract: In some examples, a method may include obtaining, from a DNA sequence, a DNA bin that includes a number of consecutive DNA elements equal to a bin length parameter and constructing sentences from the DNA bin to form a constructed sentence set that includes a number of sentences equal to a size parameter. Each sentence of the constructed sentence set may be constructed by partitioning the DNA bin into words, each word comprising a number of DNA elements equal to the size parameter. Each sentence of the constructed sentence set may include overlapping DNA elements with other sentences of the constructed sentence set and may start with a different DNA element of the DNA bin. The method may further include using the constructed sentence set to train a classifier and determining a DNA classification for an unclassified DNA subsequence through the classifier trained using the constructed sentence set.

公开(公告)号：US20180113978A1

公开(公告)日：2018-04-26

申请号：US15298412

申请日：2016-10-20

Applicant: Hewlett Packard Enterprise Development LP

Inventor： Mehran Kafai ,  Kari Lam

IPC: G06F19/28 ,  G06F17/30 ,  G06N3/08

CPC classification number: G06F19/28 ,  G06F17/30707 ,  G06N99/005

Abstract: In some examples, a method may include obtaining, from a DNA sequence, a DNA bin that includes a number of consecutive DNA elements equal to a bin length parameter and constructing sentences from the DNA bin to form a constructed sentence set that includes a number of sentences equal to a size parameter. Each sentence of the constructed sentence set may be constructed by partitioning the DNA bin into words, each word comprising a number of DNA elements equal to the size parameter. Each sentence of the constructed sentence set may include overlapping DNA elements with other sentences of the constructed sentence set and may start with a different DNA element of the DNA bin. The method may further include using the constructed sentence set to train a classifier and determining a DNA classification for an unclassified DNA subsequence through the classifier trained using the constructed sentence set.