Migration decision window selection based on hotspot characteristics

Patent metadata

Field	Value
Publication number	US-2016004473-A1
Application number	US-201414324276-A
Country	US
Kind code	A1
Filing date	Jul 7, 2014
Priority date	Jul 7, 2014
Publication date	Jan 7, 2016
Grant date	—

Abstract

Methods and arrangements for selecting a migration decision window for hotspots in a multi-tier enterprise storage system. Aspects include collecting usage statistics for data stored in the multi-tier enterprise storage system, identifying hotspots from data stored in the multi-tier enterprise storage system based on the usage statistics, and determining one or more characteristics of the identified hotspots. Aspects further include calculating an average lifespan of the identified hotspots based on the one or more characteristics of the identified hotspots and selecting the migration decision window based on the average lifespan of the identified hotspots and the one or more characteristics of the identified hotspots.

First claim

1 . A method for creating a classification rule by combining classifiers comprising:
receiving N training samples d, wherein each of the N training samples d includes a label l; receiving T classifiers C; initializing a first random weight vector α for the N training samples d; initializing a second random weight vector β for the T classifiers C; creating, by a processor, the classification rule by identifying a combination of one or more of the T classifiers C that best approximates the label l for each of the N training samples d based on the first random weight vector and the second random weight vector β.

2 . The method of claim 1 , identifying the combination of one or more of the T classifiers C that best approximates the label l for each of the N training samples d is performed by minimizing

∑

i
=
1

N





(

α
i



C


(

d
i

)

t


β

-

l
i

)

2

.

3 . The method of claim 1 , wherein each of the labels l consists of a −1 or a +1 value.

4 . The method of claim 1 , further comprising determining a similarity D of a first classifiers C p and a second classifier C q by taking a dot product of C p and C q

5 . The method of claim 4 , wherein identifying the combination of one or more of the T classifiers C that best approximates the label l further comprises minimizing

∑

i
=
1

N





∑

j
=
1

N





D


(

C


(

d
i

)

,

C


(

d
j

)

)



(

α
i

-

α
j

)

2

.

6 . The method of claim 1 , wherein e(C t ) is a set of output from applying the t th classifier C t to all the training samples d.

7 . The method of claim 6 , further comprising determining a similarity D of a first sample e(C p ) and a second sample e(C q ) by taking a dot product of e(C p ) and e(C q ).

8 . The method of claim 7 , wherein identifying the combination of one or more of the T classifiers C that best approximates the label l further comprises minimizing

∑

p
=
1

T





∑

q
=
1

T





D


(

e


(

C
p

)

,

e


(

C
q

)

)



(

β
p

-

β
q

)

2

.

9 . A computer program product for creating a classification rule by combining classifiers, the computer program produc

Assignees

• IBM

Inventors

• Chiu Lawrence Y
• Seshadri Sangeetha
• Zhou Yang

Classifications

• G06F3/0608

  Saving storage space on storage systems · CPC title

• G06F3/067

  Distributed or networked storage systems, e.g. storage area networks [SAN], network attached storage [NAS] · CPC title

• G06F3/0649 Primary

  Lifecycle management · CPC title

• G06F3/061

  Improving I/O performance · CPC title

• G06F3/0685

  Hybrid storage combining heterogeneous device types, e.g. hierarchical storage, hybrid arrays · CPC title