CN109976916A - A kind of cloud resource demand determination method and system - Google Patents
A kind of cloud resource demand determination method and system Download PDFInfo
- Publication number
- CN109976916A CN109976916A CN201910273059.0A CN201910273059A CN109976916A CN 109976916 A CN109976916 A CN 109976916A CN 201910273059 A CN201910273059 A CN 201910273059A CN 109976916 A CN109976916 A CN 109976916A
- Authority
- CN
- China
- Prior art keywords
- sample
- cloud resource
- software
- resource demand
- random forest
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Granted
Links
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/24—Classification techniques
- G06F18/243—Classification techniques relating to the number of classes
- G06F18/24323—Tree-organised classifiers
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F9/00—Arrangements for program control, e.g. control units
- G06F9/06—Arrangements for program control, e.g. control units using stored programs, i.e. using an internal store of processing equipment to receive or retain programs
- G06F9/46—Multiprogramming arrangements
- G06F9/50—Allocation of resources, e.g. of the central processing unit [CPU]
- G06F9/5005—Allocation of resources, e.g. of the central processing unit [CPU] to service a request
- G06F9/5027—Allocation of resources, e.g. of the central processing unit [CPU] to service a request the resource being a machine, e.g. CPUs, Servers, Terminals
- G06F9/5055—Allocation of resources, e.g. of the central processing unit [CPU] to service a request the resource being a machine, e.g. CPUs, Servers, Terminals considering software capabilities, i.e. software resources associated or available to the machine
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Data Mining & Analysis (AREA)
- General Engineering & Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Physics & Mathematics (AREA)
- Software Systems (AREA)
- Life Sciences & Earth Sciences (AREA)
- Artificial Intelligence (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Bioinformatics & Computational Biology (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Evolutionary Biology (AREA)
- Evolutionary Computation (AREA)
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
Abstract
The present invention provides a kind of cloud resource demand determination method, belongs to field of cloud computer technology, can at least partly solve the problems, such as method and software in the prior art without judging the cloud resource demand of software.A kind of cloud resource demand determination method of the invention, comprising: choose at least two specific softwares in the software that cloud resource platform has been disposed as sample;According to sample and at least two special characteristics of sample, at least two decision trees are formed, to form random forest, wherein special characteristic indicates feature relevant to the cloud resource demand of software;Using random forest assessment to the cloud resource demand of deployment software in cloud resource platform.
Description
Technical field
The invention belongs to field of cloud computer technology, and in particular to a kind of cloud resource demand determination method and system.
Background technique
Cloud computing technology is using increasingly extensive, and information technology resources are encapsulated as cloud resource by cloud management platform, when software exists
When cloud resource Platform deployment, for the software distribution cloud resource and charging is carried out.
The scheme that existing cloud resource demand determines is by the micro-judgment of operation maintenance personnel to each software distribution cloud
Resource, this method lack the believable data reference of science, only subjective judgement, are easy to appear resource size and software is practical
The unmatched situation of demand.
Summary of the invention
The present invention at least partly solves the method for cloud resource demand and asking for software without judging software in the prior art
Topic provides a kind of cloud resource demand determination method and system.
Solving technical solution used by present invention problem is a kind of cloud resource demand determination method, comprising:
At least two specific softwares in the software that cloud resource platform has been disposed are chosen as sample;
According to the sample and at least two special characteristics of the sample, at least two decision trees are formed, to be formed
Random forest, wherein the special characteristic indicates feature relevant to the cloud resource demand of software;
Using random forest assessment to the cloud resource demand of deployment software in cloud resource platform.
It may further be preferable that the specific software chosen in the software that cloud resource platform has been disposed is as sample packet
Include: multiple selected properties of deployment software are respectively to the contributive rate of the cloud resource demand of the software in statistics cloud resource platform;
It is obtained according to the preset weighted value of the contributive rate of each selected properties and each selected properties each described soft
The resource matched score of part, the resource matched score indicate the cloud resource demand of software and the matching degree of practical cloud resource;Choosing
At least two softwares of the fixed resource matched highest scoring are as the specific software.
It may further be preferable that the step of obtaining the special characteristic includes: the cloud resource demand being calculated with software
The unstability index of relevant all features;The smallest at least two features of unstability index are chosen as the specific spy
Sign.
It may further be preferable that the calculation formula of the unstability index is as follows:
Wherein, f indicates feature relevant to the cloud resource demand of software, and n indicates the n attribute of this feature, pkIndicate kth
The probability that attribute occurs.
It may further be preferable that it is described according to the sample and at least two special characteristics of the sample, it is formed extremely
Few two decision trees, include: to select the first training sample at random from the sample to form random forest;Using putting back at random
Method m sample is chosen in first training sample;According to m selected sample and randomly select j specific spies
Sign forms the first decision tree;According to m selected sample and k special characteristics are randomly selected, form the second decision tree, j is not
Equal to k;First decision tree and second decision tree is set to form the first random forest.
It may further be preferable that it is described according to the sample and at least two special characteristics of the sample, it is formed extremely
Few two decision trees, to form random forest further include: select the second training sample at random from the sample;Using putting at random
The method returned chooses m sample in second training sample;According to m selected sample and randomly select p it is specific
Feature forms third decision tree;According to m selected sample and q special characteristics are randomly selected, form the 4th decision tree, p
Not equal to q;The third decision tree and the 4th decision tree is set to form the second random forest.
It may further be preferable that the quantity of software is respectively greater than first training sample and described second in the sample
The quantity of software in training sample is the first verifying sample, the sample in addition to first training sample in the sample
It is the second verifying sample in addition to second training sample in this;It is described to form first random forest and second at random
After forest further include: by the special characteristic of the first verifying sample, calculate the serious forgiveness of first random forest;It is logical
The special characteristic for crossing the second verifying sample, calculates the serious forgiveness of second random forest;It is random gloomy to choose described first
The formation resource requirement that serious forgiveness is low in woods and the second random forest prejudges model.
It may further be preferable that described provided in cloud resource platform to the cloud of deployment software using random forest assessment
Source demand includes: that each feature to deployment software is transported to the resource requirement anticipation model, to be disposed to assess
The cloud resource demand of software.
Solving technical solution used by present invention problem is a kind of cloud resource demand decision-making system, comprising:
Message processing module, for choosing at least two specific softwares in the software that cloud resource platform has been disposed as sample
This;
Model construction module is formed at least at least two special characteristics according to the sample and the sample
Two decision trees, to form random forest, wherein the special characteristic indicates feature relevant to the cloud resource demand of software;
Demand determination module, for being assessed in cloud resource platform using the random forest to the cloud resource of deployment software
Demand.
It may further be preferable that the cloud resource demand decision-making system further include: Feature Selection module, for be calculated with
The unstability index of the relevant all features of the cloud resource demand of software and selection the smallest at least two institutes of unstability index
Feature is stated as the special characteristic.
In the cloud resource demand determination method of the present embodiment, according to the software disposed in cloud resource platform and its specific
Feature forms the demand to cloud resource that can be evaluated whether software, so as to the very convenient money for efficiently judging non-deployment software
Whether source demand application is reasonable, i.e., under the premise of meeting resource needed for software, improves the utilization rate of cloud resource.
Detailed description of the invention
Fig. 1 is a kind of flow chart of cloud resource demand determination method of the embodiment of the present invention;
Fig. 2 is a kind of flow chart of cloud resource demand determination method of the embodiment of the present invention;
Fig. 3 is a kind of structural schematic diagram of cloud resource demand decision-making system of the embodiment of the present invention.
Specific embodiment
Technical solution in order to enable those skilled in the art to better understand the present invention, with reference to the accompanying drawing and specific embodiment party
Present invention is further described in detail for formula.
Embodiment 1:
As shown in Figure 1, the present embodiment provides a kind of cloud resource demand determination methods, comprising:
At least two specific softwares in software (i.e. software systems) that S11, selection cloud resource platform have been disposed are as sample
This.
Wherein, software deployed in cloud resource platform refers to the software that cloud resource platform has carried, that is, distribute to
The resource of the software is known.Specific software indicate the cloud resource amount that is assigned in all deployment softwares and itself
The more matched software of demand.
S12, according at least two special characteristics of sample and sample, form at least two decision trees, it is random to be formed
Forest, wherein special characteristic indicates feature relevant to the cloud resource demand of software.
Wherein, that is to say, that sample and different special characteristics form different decision trees, then are formed by multiple decision trees
One random forest.
S13, it is assessed in cloud resource platform using random forest to the cloud resource demand of deployment software.
Wherein, the software not carried also to deployment software expression cloud resource platform, i.e. cloud resource platform do not give the software also
Distribute cloud resource.It can be evaluated whether to obtain demand of the software to cloud resource by established random forest.
In the cloud resource demand determination method of the present embodiment, according to the software disposed in cloud resource platform and its specific
Feature forms the demand to cloud resource that can be evaluated whether software, so as to the very convenient money for efficiently judging non-deployment software
Whether source demand application is reasonable, i.e., under the premise of meeting resource needed for software, improves the utilization rate of cloud resource.
Embodiment 2:
As shown in Fig. 2, the present embodiment provides a kind of cloud resource demand determination methods, comprising:
At least two specific softwares in software that S21, selection cloud resource platform have been disposed are as sample.
Wherein, software deployed in cloud resource platform refers to the software that cloud resource platform has carried, that is, distribute to
The resource of the software is known.Specific software indicate the cloud resource amount that is assigned in all deployment softwares and itself
The more matched software of demand.
Specifically, the specific software chosen in the software that cloud resource platform has been disposed includes: as sample
Multiple selected properties of deployment software are respectively to the cloud resource demand of the software in S211, statistics cloud resource platform
Contributive rate.
Wherein, which may include the service quality (SLA) and resource utilization to software users.Specifically,
Intend considering that user's average retardation, moon failure rate, troubleshooting are averaged the indexs such as duration to the service quality (SLA) of software users;Resource
Utilization rate is quasi- to consider that central processing unit (CPU) utilization rate, memory usage, input/output (I/O) service condition, disk utilize
The indexs such as rate.
S212, obtained according to the contributive rate of each selected properties and the preset weighted value of each selected properties it is each soft
The resource matched score of part, resource matched score indicate the cloud resource demand of software and the matching degree of practical cloud resource.
Wherein, the preset weighted value of selected properties can be the weight column for each selected properties being empirically derived
Table, for example, (ω1, ω2...ωn), ωnIndicate n-th of selected properties weighted value.
The calculation formula of the resource matched score of each software is as follows:
C=ω1×n1+ω2×n2+...+ωn×nn, wherein ωnIndicate n-th of selected properties weighted value, nnIndicate n-th
Contributive rate of a selected properties to the cloud resource demand of software.
That is, resource matched score is higher, then the cloud resource demand of the software itself and its actual cloud resource
It is higher with spending.
S213, at least two softwares of resource matched highest scoring are selected as specific software.
Wherein, that is to say, that the software for selecting resource matched score high is as specific software.
For example, if desired selecting 10 specific softwares as sample, then before selecting the descending ranking of resource matched score
10 conduct specific software.
S22, according at least two special characteristics of sample and sample, form at least two decision trees, it is random to be formed
Forest, wherein special characteristic indicates feature relevant to the cloud resource demand of software.
Wherein, that is to say, that sample and different special characteristics form different decision trees, then are formed by multiple decision trees
One random forest.
Specifically, the feature of the quasi- choosing of calculating demand for the resource requirement anticipation model that special characteristic is formed after including and
The feature of the quasi- choosing of storage demand.The feature that resource requirement prejudges the quasi- choosing of calculating demand of model includes prospective users amount, Yong Huzeng
Long rate, peak value concurrent user amount, the configuration of dsc data amount, virtual machine number, load balancing etc.;The storage of resource requirement anticipation model
The feature of the quasi- choosing of demand includes that service scripts type, the average daily growth rate of service scripts, business datum table size, user data table are big
Small, index quantity, monthly user's growth rate.
Wherein, the step of obtaining special characteristic include:
The unstability index of all features relevant to the cloud resource demand of software is calculated;Choose unstability index most
Small at least two features are as special characteristic.
Specifically, the calculation formula of unstability index is as follows:
Wherein, f indicates feature relevant to the cloud resource demand of software, and n indicates the n attribute of this feature, PkIndicate kth
The probability that attribute occurs.
It should be noted that in a practical situation, preferentially selecting the smallest 10 features of unstability index as specific spy
Sign.
Wherein, according to sample and at least two special characteristics of sample, at least two decision trees are formed, it is random to be formed
Forest includes:
S221, the first training sample is selected at random from sample;
Wherein, that is to say, that the first verifying sample that sample is randomly divided into the first training sample and is hereinafter mentioned.
Specifically, the quantity of the first training sample can account for the 80% of all samples, and the quantity of the first verifying sample can account for it is all
The 20% of sample.
S222, using there is the method put back to choose m sample in the first training sample at random.
S223, according to m selected sample and j special characteristics are randomly selected, forms the first decision tree.
S224, according to m selected sample and k special characteristics are randomly selected, forms the second decision tree, j is not equal to
k。
S225, the first decision tree and the second decision tree is made to form the first random forest.
It should be noted that in a practical situation, if the quantity of special characteristic is 10, according in the first training sample
In selected m sample and special characteristic can form 10 decision trees, for example, according to m selected sample and at random
1 special characteristic of choosing forms a decision tree, the 2 special characteristics formation one selected according to m selected sample and at random
A decision tree etc..A random forest finally is formed with this 10 decision trees, which can more accurately assess in cloud
To the cloud resource demand of deployment software in resource platform.
S226, the second training sample is selected at random from sample.
Wherein, that is to say, that the second verifying sample that sample is randomly divided into the second training sample and is hereinafter mentioned.
Specifically, the quantity of the second training sample can account for the 80% of all samples, and the quantity of the second verifying sample can account for it is all
The 20% of sample.
S227, using there is the method put back to choose m sample in the second training sample at random.
S228, according to m selected sample and p special characteristics are randomly selected, forms third decision tree.
S229, according to m selected sample and q special characteristics are randomly selected, forms the 4th decision tree, p is not equal to
q。
S2210, third decision tree and the 4th decision tree is made to form the second random forest.
Wherein, it should be noted that in a practical situation, if the quantity of special characteristic is 10, according in the second instruction
10 decision trees can be formed by practicing m sample and special characteristic selected in sample, for example, according to m selected sample with
And 1 special characteristic selecting at random forms a decision tree, 2 special characteristics selected according to m selected sample and at random
Form decision tree etc..A random forest finally is formed with this 10 decision trees, which can more accurately comment
Estimate the cloud resource demand in cloud resource platform to deployment software.
Preferably, the quantity of software is respectively greater than the number of software in the first training sample and the second training sample in sample
Amount, is the first verifying sample in addition to the first training sample in sample, is second in addition to the second training sample in sample
Verify sample.
S2211, pass through the special characteristic of the first verifying sample, the serious forgiveness of the first random forest of calculating;
S2212, by second verifying sample special characteristic, calculate the second random forest serious forgiveness;
S2213, the low formation resource requirement anticipation model of serious forgiveness in the first random forest and the second random forest is chosen.
It should be noted that sample can be carried out to n times is randomly divided into training sample and verifying sample, in this way according to each
It is a random gloomy to form n for the secondary training sample got and verifying sample (such as step S221-S225 or step S226-S2210)
Woods.Software when due to each sub-distribution in training sample is not identical or not exclusively same, and the random forest formed is not yet
Together.
Then the serious forgiveness (N times of cross-validation method) of n random forest can be calculated according to step S2211 to get appearance is arrived
Error rate list (q1, q2…qN), qNIndicate the serious forgiveness of n-th of random forest;And according to step S2213 to select n random gloomy
Serious forgiveness is one the smallest in woods, selects the special characteristic in the random forest as preferred feature, to form final money
Source demand prejudges model.
S23, it is assessed in cloud resource platform using random forest to the cloud resource demand of deployment software.
Wherein, the software not carried also to deployment software expression cloud resource platform, cloud resource platform is not also to the software point
With cloud resource.It can be evaluated whether to obtain demand of the software to cloud resource by established random forest.
Specifically, including: using cloud resource demand of the random forest assessment in cloud resource platform to deployment software
Each feature to deployment software is transported into resource requirement anticipation model, to assess the cloud resource of software to be disposed
Demand.
In the cloud resource demand determination method of the present embodiment, according to the software disposed in cloud resource platform and its specific
Feature forms the demand to cloud resource that can be evaluated whether software, so as to the very convenient money for efficiently judging non-deployment software
Whether source demand application is reasonable, i.e., under the premise of meeting resource needed for software, improves the utilization rate of cloud resource.
Embodiment 3:
As shown in figure 3, the present embodiment provides a kind of cloud resource demand decision-making systems, comprising:
Message processing module, for choosing at least two specific softwares in the software that cloud resource platform has been disposed as sample
This;
Model construction module forms at least two decisions at least two special characteristics according to sample and sample
Tree, to form random forest, wherein special characteristic indicates feature relevant to the cloud resource demand of software;
Demand determination module, for being needed in cloud resource platform to the cloud resource of deployment software using random forest assessment
It asks.
Preferably, the cloud resource demand decision-making system further include:
Feature Selection module, for the unstability index of all features relevant to the cloud resource demand of software to be calculated
And the smallest at least two features of unstability index are chosen as special characteristic.
It should be noted that herein, relational terms such as first and second and the like are used merely to a reality
Body or operation are distinguished with another entity or operation, are deposited without necessarily requiring or implying between these entities or operation
In any actual relationship or order or sequence.Moreover, the terms "include", "comprise" or its any other variant are intended to
Non-exclusive inclusion, so that the process, method, article or equipment including a series of elements is not only wanted including those
Element, but also including other elements that are not explicitly listed, or further include for this process, method, article or equipment
Intrinsic element.In the absence of more restrictions, the element limited by sentence "including a ...", it is not excluded that
There is also other identical elements in process, method, article or equipment including element.
It is as described above according to the embodiment of the present invention, these embodiments details all there is no detailed descriptionthe, also not
Limiting the invention is only the specific embodiment.Obviously, as described above, can make many modifications and variations.This explanation
These embodiments are chosen and specifically described to book, is principle and practical application in order to better explain the present invention, thus belonging to making
Technical field technical staff can be used using modification of the invention and on the basis of the present invention well.The present invention is only by right
The limitation of claim and its full scope and equivalent.
Claims (10)
1. a kind of cloud resource demand determination method characterized by comprising
At least two specific softwares in the software that cloud resource platform has been disposed are chosen as sample;
According to the sample and at least two special characteristics of the sample, at least two decision trees are formed, it is random to be formed
Forest, wherein the special characteristic indicates feature relevant to the cloud resource demand of software;
Using random forest assessment to the cloud resource demand of deployment software in cloud resource platform.
2. cloud resource demand determination method according to claim 1, which is characterized in that selection cloud resource platform portion
Specific software in the software of administration includes: as sample
Multiple selected properties of deployment software in cloud resource platform are counted respectively to the contributive rate of the cloud resource demand of the software;
Each institute is obtained according to the preset weighted value of the contributive rate of each selected properties and each selected properties
The resource matched score of software is stated, the resource matched score indicates the matching of the cloud resource demand and practical cloud resource of software
Degree;
At least two softwares of the resource matched highest scoring are selected as the specific software.
3. cloud resource demand determination method according to claim 1, which is characterized in that the step of obtaining the special characteristic
Include:
The unstability index of all features relevant to the cloud resource demand of software is calculated;
The smallest at least two features of unstability index are chosen as the special characteristic.
4. cloud resource demand determination method according to claim 3, which is characterized in that the calculating of the unstability index is public
Formula is as follows:
Wherein, f indicates feature relevant to the cloud resource demand of software, and n indicates the n attribute of this feature, pkIndicate kth kind
Property occur probability.
5. cloud resource demand determination method according to claim 1, which is characterized in that described according to the sample and institute
At least two special characteristics of sample are stated, at least two decision trees is formed, includes: to form random forest
The first training sample is selected at random from the sample;
Using there is the method put back to choose m sample in first training sample at random;
According to m selected sample and j special characteristics are randomly selected, form the first decision tree;
According to m selected sample and k special characteristics are randomly selected, form the second decision tree, j is not equal to k;
First decision tree and second decision tree is set to form the first random forest.
6. cloud resource demand determination method according to claim 5, which is characterized in that described according to the sample and institute
At least two special characteristics of sample are stated, at least two decision trees are formed, to form random forest further include:
The second training sample is selected at random from the sample;
Using there is the method put back to choose m sample in second training sample at random;
According to m selected sample and p special characteristics are randomly selected, form third decision tree;
According to m selected sample and q special characteristics are randomly selected, form the 4th decision tree, p is not equal to q;
The third decision tree and the 4th decision tree is set to form the second random forest.
7. cloud resource demand determination method according to claim 6, which is characterized in that the quantity of software point in the sample
Not great Yu in first training sample and second training sample software quantity, except first training in the sample
It is the first verifying sample except sample, is the second verifying sample in addition to second training sample in the sample;
It is described to be formed after first random forest and the second random forest further include:
By the special characteristic of the first verifying sample, the serious forgiveness of first random forest is calculated;
By the special characteristic of the second verifying sample, the serious forgiveness of second random forest is calculated;
Choose the formation resource requirement anticipation model that serious forgiveness is low in first random forest and the second random forest.
8. cloud resource demand determination method according to claim 7, which is characterized in that described to be commented using the random forest
Estimating the cloud resource demand in cloud resource platform to deployment software includes:
Each feature to deployment software is transported into the resource requirement anticipation model, to assess the cloud of software to be disposed
Resource requirement.
9. a kind of cloud resource demand decision-making system characterized by comprising
Message processing module, for choosing at least two specific softwares in the software that cloud resource platform has been disposed as sample;
Model construction module forms at least two at least two special characteristics according to the sample and the sample
Decision tree, to form random forest, wherein the special characteristic indicates feature relevant to the cloud resource demand of software;
Demand determination module, for being needed in cloud resource platform to the cloud resource of deployment software using random forest assessment
It asks.
10. cloud resource demand decision-making system according to claim 9, which is characterized in that further include:
Feature Selection module, for be calculated all features relevant to the cloud resource demand of software unstability index and
The smallest at least two features of unstability index are chosen as the special characteristic.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201910273059.0A CN109976916B (en) | 2019-04-04 | 2019-04-04 | Cloud resource demand judgment method and system |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201910273059.0A CN109976916B (en) | 2019-04-04 | 2019-04-04 | Cloud resource demand judgment method and system |
Publications (2)
Publication Number | Publication Date |
---|---|
CN109976916A true CN109976916A (en) | 2019-07-05 |
CN109976916B CN109976916B (en) | 2021-05-11 |
Family
ID=67083192
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201910273059.0A Active CN109976916B (en) | 2019-04-04 | 2019-04-04 | Cloud resource demand judgment method and system |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN109976916B (en) |
Citations (12)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
WO2015062209A1 (en) * | 2013-10-29 | 2015-05-07 | 华为技术有限公司 | Visualized optimization processing method and device for random forest classification model |
CN104798043A (en) * | 2014-06-27 | 2015-07-22 | 华为技术有限公司 | Data processing method and computer system |
CN105101217A (en) * | 2014-05-15 | 2015-11-25 | 中国移动通信集团安徽有限公司 | Method and equipment for configuring network resources |
CN105227369A (en) * | 2015-10-19 | 2016-01-06 | 南京华苏科技股份有限公司 | Based on the mobile Apps of mass-rent pattern to the analytical method of the Wi-Fi utilization of resources |
US20160321549A1 (en) * | 2015-04-29 | 2016-11-03 | Facebook, Inc. | Boosted decision trees for evaluating feature vectors |
CN106897109A (en) * | 2017-02-13 | 2017-06-27 | 云南大学 | Based on the virtual machine performance Forecasting Methodology that random forest is returned |
CN107404523A (en) * | 2017-07-21 | 2017-11-28 | 中国石油大学(华东) | Cloud platform adaptive resource dispatches system and method |
CN107729555A (en) * | 2017-11-07 | 2018-02-23 | 太原理工大学 | A kind of magnanimity big data Distributed Predictive method and system |
CN108182115A (en) * | 2017-12-28 | 2018-06-19 | 福州大学 | A kind of virtual machine load-balancing method under cloud environment |
CN108710999A (en) * | 2018-05-03 | 2018-10-26 | 上海电机学院 | The confidence level automatic evaluation method of shared resource under a kind of environment based on big data |
CN109284871A (en) * | 2018-09-30 | 2019-01-29 | 北京金山云网络技术有限公司 | Resource adjusting method, device and cloud platform |
US20190034242A1 (en) * | 2017-07-26 | 2019-01-31 | Bank Of America Corporation | Multi-system event response calculator and resource allocator |
-
2019
- 2019-04-04 CN CN201910273059.0A patent/CN109976916B/en active Active
Patent Citations (12)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
WO2015062209A1 (en) * | 2013-10-29 | 2015-05-07 | 华为技术有限公司 | Visualized optimization processing method and device for random forest classification model |
CN105101217A (en) * | 2014-05-15 | 2015-11-25 | 中国移动通信集团安徽有限公司 | Method and equipment for configuring network resources |
CN104798043A (en) * | 2014-06-27 | 2015-07-22 | 华为技术有限公司 | Data processing method and computer system |
US20160321549A1 (en) * | 2015-04-29 | 2016-11-03 | Facebook, Inc. | Boosted decision trees for evaluating feature vectors |
CN105227369A (en) * | 2015-10-19 | 2016-01-06 | 南京华苏科技股份有限公司 | Based on the mobile Apps of mass-rent pattern to the analytical method of the Wi-Fi utilization of resources |
CN106897109A (en) * | 2017-02-13 | 2017-06-27 | 云南大学 | Based on the virtual machine performance Forecasting Methodology that random forest is returned |
CN107404523A (en) * | 2017-07-21 | 2017-11-28 | 中国石油大学(华东) | Cloud platform adaptive resource dispatches system and method |
US20190034242A1 (en) * | 2017-07-26 | 2019-01-31 | Bank Of America Corporation | Multi-system event response calculator and resource allocator |
CN107729555A (en) * | 2017-11-07 | 2018-02-23 | 太原理工大学 | A kind of magnanimity big data Distributed Predictive method and system |
CN108182115A (en) * | 2017-12-28 | 2018-06-19 | 福州大学 | A kind of virtual machine load-balancing method under cloud environment |
CN108710999A (en) * | 2018-05-03 | 2018-10-26 | 上海电机学院 | The confidence level automatic evaluation method of shared resource under a kind of environment based on big data |
CN109284871A (en) * | 2018-09-30 | 2019-01-29 | 北京金山云网络技术有限公司 | Resource adjusting method, device and cloud platform |
Non-Patent Citations (2)
Title |
---|
DEEPAK JAIN 等: "Optimization of resource and task scheduling in cloud using random forest", 《2017 INTERNATIONAL CONFERENCE ON ADVANCES IN COMPUTING, COMMUNICATION AND CONTROL (ICAC3)》 * |
朱元昌 等: ""IaaS模式云训练资源预测-调度方法"", 《系统工程与电子技术》 * |
Also Published As
Publication number | Publication date |
---|---|
CN109976916B (en) | 2021-05-11 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
US11233765B2 (en) | Systems and methods for relaying messages in a communications system | |
US8250008B1 (en) | Decision tree refinement | |
US20180247265A1 (en) | Task grouping method and apparatus, electronic device, and computer storage medium | |
Rowe et al. | Automated social hierarchy detection through email network analysis | |
US20190089610A1 (en) | Parallel computational framework and application server for determining path connectivity | |
CN110377804A (en) | Method for pushing, device, system and the storage medium of training course data | |
CN108399453A (en) | A kind of Electric Power Customer Credit Rank Appraisal method and apparatus | |
CN109558989A (en) | Queuing time prediction technique, device, equipment and computer readable storage medium | |
CN109783227A (en) | Method for allocating tasks, device, system and computer readable storage medium | |
CN106445929A (en) | Activeness based user ranking method and system | |
US20140164364A1 (en) | System and method for event-driven prioritization | |
CN106570763A (en) | User influence evaluation method and system | |
CN110457577A (en) | Data processing method, device, equipment and computer storage medium | |
CN110557345B (en) | Power communication network resource allocation method | |
CN105843931A (en) | Classification method and device | |
CN104461736B (en) | Resource allocation and searching method, resource allocation and search system and Cloud Server | |
CN109976916A (en) | A kind of cloud resource demand determination method and system | |
CN105786622B (en) | A kind of node selecting method calculated under cloud environment for real-time collaborative | |
CN102737044B (en) | Method and device for releasing webpage information | |
CN107277143A (en) | A kind of resource matched management method and device | |
CN112712210B (en) | Sudden topic transmission scale prediction method, system, processing terminal and medium | |
CN110348807A (en) | A kind of information processing method and relevant apparatus | |
CN108347466A (en) | A kind of date storage method and device of cloud storage system | |
Phythian et al. | Progress in electronic service delivery by English District Councils | |
CN115293275A (en) | Data identification method and device, electronic equipment and storage medium |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
PB01 | Publication | ||
PB01 | Publication | ||
SE01 | Entry into force of request for substantive examination | ||
SE01 | Entry into force of request for substantive examination | ||
GR01 | Patent grant | ||
GR01 | Patent grant |