CN108764474A - A kind of neuron parameter fitness method - Google Patents
A kind of neuron parameter fitness method Download PDFInfo
- Publication number
- CN108764474A CN108764474A CN201810517988.7A CN201810517988A CN108764474A CN 108764474 A CN108764474 A CN 108764474A CN 201810517988 A CN201810517988 A CN 201810517988A CN 108764474 A CN108764474 A CN 108764474A
- Authority
- CN
- China
- Prior art keywords
- individual
- parameter
- value
- population
- action potential
- 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.)
- Pending
Links
- 210000002569 neuron Anatomy 0.000 title claims abstract description 27
- 238000000034 method Methods 0.000 title claims abstract description 24
- 238000004422 calculation algorithm Methods 0.000 claims abstract description 22
- 102000004310 Ion Channels Human genes 0.000 claims abstract description 18
- 230000002068 genetic effect Effects 0.000 claims abstract description 18
- 230000008569 process Effects 0.000 claims abstract description 7
- 230000036982 action potential Effects 0.000 claims description 25
- 108090000862 Ion Channels Proteins 0.000 claims description 16
- 238000009826 distribution Methods 0.000 claims description 6
- 102000004257 Potassium Channel Human genes 0.000 claims description 4
- 108010052164 Sodium Channels Proteins 0.000 claims description 4
- 102000018674 Sodium Channels Human genes 0.000 claims description 4
- 150000002500 ions Chemical class 0.000 claims description 4
- 230000035772 mutation Effects 0.000 claims description 4
- 108020001213 potassium channel Proteins 0.000 claims description 4
- 108090000623 proteins and genes Proteins 0.000 claims description 3
- 238000006467 substitution reaction Methods 0.000 claims description 3
- 241000208340 Araliaceae Species 0.000 claims description 2
- 108090000312 Calcium Channels Proteins 0.000 claims description 2
- 102000003922 Calcium Channels Human genes 0.000 claims description 2
- 108091006146 Channels Proteins 0.000 claims description 2
- 235000005035 Panax pseudoginseng ssp. pseudoginseng Nutrition 0.000 claims description 2
- 235000003140 Panax quinquefolius Nutrition 0.000 claims description 2
- 235000008434 ginseng Nutrition 0.000 claims description 2
- 238000012421 spiking Methods 0.000 claims description 2
- 238000010353 genetic engineering Methods 0.000 claims 1
- 230000000694 effects Effects 0.000 abstract description 4
- 230000001537 neural effect Effects 0.000 abstract description 2
- HPTJABJPZMULFH-UHFFFAOYSA-N 12-[(Cyclohexylcarbamoyl)amino]dodecanoic acid Chemical compound OC(=O)CCCCCCCCCCCNC(=O)NC1CCCCC1 HPTJABJPZMULFH-UHFFFAOYSA-N 0.000 abstract 1
- 238000004364 calculation method Methods 0.000 description 3
- 238000010586 diagram Methods 0.000 description 2
- 230000005611 electricity Effects 0.000 description 2
- 238000005457 optimization Methods 0.000 description 2
- 238000012800 visualization Methods 0.000 description 2
- 241001269238 Data Species 0.000 description 1
- 230000009471 action Effects 0.000 description 1
- 230000006978 adaptation Effects 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 210000004027 cell Anatomy 0.000 description 1
- 238000000205 computational method Methods 0.000 description 1
- 230000008676 import Effects 0.000 description 1
- 230000037361 pathway Effects 0.000 description 1
- 238000002360 preparation method Methods 0.000 description 1
- 230000009467 reduction Effects 0.000 description 1
- 238000004088 simulation Methods 0.000 description 1
- 241000894007 species Species 0.000 description 1
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/08—Learning methods
- G06N3/082—Learning methods modifying the architecture, e.g. adding, deleting or silencing nodes or connections
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/06—Physical realisation, i.e. hardware implementation of neural networks, neurons or parts of neurons
- G06N3/061—Physical realisation, i.e. hardware implementation of neural networks, neurons or parts of neurons using biological neurons, e.g. biological neurons connected to an integrated circuit
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/12—Computing arrangements based on biological models using genetic models
- G06N3/126—Evolutionary algorithms, e.g. genetic algorithms or genetic programming
Landscapes
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- Health & Medical Sciences (AREA)
- Life Sciences & Earth Sciences (AREA)
- Biophysics (AREA)
- Theoretical Computer Science (AREA)
- Biomedical Technology (AREA)
- Molecular Biology (AREA)
- Computing Systems (AREA)
- Artificial Intelligence (AREA)
- General Health & Medical Sciences (AREA)
- Data Mining & Analysis (AREA)
- Computational Linguistics (AREA)
- General Engineering & Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Mathematical Physics (AREA)
- Software Systems (AREA)
- Evolutionary Computation (AREA)
- Neurology (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Bioinformatics & Computational Biology (AREA)
- Evolutionary Biology (AREA)
- Microelectronics & Electronic Packaging (AREA)
- Physiology (AREA)
- Genetics & Genomics (AREA)
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
Abstract
A kind of neuron parameter fitness method belongs to neuroscience field, improves the efficiency of neuron parameter fitting.Because restoring true Neural spike train activity in reality to be of great significance, the ion channel type that certain neuron is included can be measured in an experiment, but the electric conductivity value of ion channel is not easy to measure.The present invention accelerates the process that genetic algorithm solves neuron models conductance parameter using CUDA programmings, takes full advantage of the computing resource of GPU, and the process of algorithm solution has been visualized using MFC frames.The tool can also both test to the operation result of tool to generate pseudo- experimental data according to specified conductance parameter value and stimulated current by importing experimental data.This tool facilitates the use of experimenter.Experimental result shows that algorithm can find the value close to optimized parameter, to reappear the electric signal that experiment measures.
Description
Technical field：
The invention belongs to neuroscience fields.Be it is a kind of parallel genetic algorithm is utilized experimental data is fitted seek
The method for looking for parameter.This method can effectively improve the efficiency of neuron parameter fitting.
Background technology
For the HH models containing different kinds of ions channel, expression formula shaped likeWe can measure the ion channel that certain neuron is included in an experiment
Type, if it is known that conductance parameterValue, we can solve the above-mentioned differential equation and obtain the voltage value V at each moment, this
Sample we can go to simulate true Neural spike train activity with this equation.It is very tired how optimal parameter is looked for
Difficulty, such as in the experiment for measuring electric signal, even if if having been known for the ion channel type that neuronal cell is included very
Hardly possible simulates the discharge activities of the cell, because being usually difficult to measure the electric conductivity value in relation to ion channel in an experiment, even together
One equation, different conductance parameters, generated electric signal can be also very different.Some present neuron simulation tools
As NEURON and GENESIS also provide relevant parameter optimization tool.These softwares are programmed using CPU a bit, some are only
Very simple gui interface is provided, still have greatly improved space for speed and practicability.Therefore, this paper presents one
The neuron parameter fitness method of kind highly effective.
Invention content：
The present invention is to solve the problem of neuron conductance parameter fitting, in the case where there are many experimental data, manually
Ginseng is adjusted to become highly difficult.If carrying out Automatic parameter fitting using genetic algorithm, when individual amount is very more in population,
It is be easy to cause Calculation bottleneck, therefore this paper presents a kind of efficient computational methods.
Initial population is generated according to the parameter initialization population of input first, what each individual in population actually represented
It is exactly one group of parameter, also needs one group of true peak hour point data of input, this group of data can be by true biological experiment
It measures record to obtain, can also be generated by algorithm.Each individual in population can generate one group of point after substituting into HH equations
It is right can to calculate each individual according to this group of peak hour point data and the true spiking data for starting to input for peak time point data
The fitness value answered.Particularly, in order to solve when in population individual amount it is many and caused by calculate the bottleneck of fitness value, this
The process that each individual calculates fitness value is put into GPU cores and carries out by invention.Then it can be lost according to fitness value
Pass operation, final output optimum individual, i.e., optimal conductance parameter.It is as follows：
(1) one group of true peak hour point data is inputted, this group of data can be measured by true biological experiment to be remembered
Record obtains, can also the pseudo- experimental data method generation of generation provided by the present invention.
Wherein generate pseudo- experimental data comprises the concrete steps that the value for first giving conductance parameter, reselection need the ion being fitted
Then pathway species substitute into following HH equations and obtain data.
Wherein CMIndicate that capacitance, V refer to voltage value,It is electric conductivity value, m, n, h is the differential equation about t, and I refers to
The current stimulated current for giving neuronal cell, includes a sodium channel and a potassium channel in this equation, the m in above formula,
N, h reflect the opening and closing dynamic process of these three ion channels respectively.αn, βnDeng the rate for indicating associated ion channel
Constant.
(2) it uses paralleling genetic algorithm to carry out parameter fitting to the data in (1), is as follows：
1) real coding is carried out to conductance parameter first, x=[x1, x2, x3, x4, x5] wherein x1~x5 represents five electricity
Parameter is led, including a sodium channel, a calcium channel and three potassium channels, each group of conductance parameter represent an individual,
The initial population A for including many this individuals is generated later.
2) it selects, excellent individual is replicated according to fitness value, eliminates poor individual：
(GPU specific implementations are to select a Small Population A1 from initial population A, population A 1 is divided equally into two parts a1,
A2, compares a1, the size of individual fitness value on the corresponding position of a2, if the fitness value ratio of the individual in a2 on some position
A1 is good, with the individual of corresponding position on the individual substitution a1 in a2)
3) mixing intersects：
(mixing is carried out according to crossover probability to each of a2 individuals to intersect, i.e., to each of a2 individual c, existed at random
Two individuals n1, n2 are selected in a1, according to formula Ki=n1*weight+n2* (1-weight) executes intersection, weight 0-1
Between random value.)
4) it makes a variation：
(normal distribution variation is carried out according to mutation probability to each of a2 individuals, i.e., to each of a2 individuals KiHave
Five gene positions, i.e., the parameter value of five kinds ion channels, Ki=[x1, x2, x3, x4, x5] is then directed to gene position x1 (first
Ion channel), x1=weight, wherein weight Normal Distributions, set here the mean value of its Normal Distribution as
(x1_Low+x1_Up)/2, x1_Low indicates that the lower bound of first ion channel parameter value of neuron, x1_Up indicate neuron the
The upper bound of one ion channel parameter value)
5) population exchanges：
(from some individuals being introduced into a1 in some individual substitution population As, some individuals are introduced from A, are replaced in a1
Some individuals.)
6) return of population A 1 is obtained into new population A：
7) 2-6 is repeated, the maximum genetic algebra until reaching setting returns the result.
The calculating of wherein fitness value uses following method：
1. it is M, wherein m to define an action potential sequence firstiIt indicates to reach i-th of Voltage Peak in action potential sequence
The time point of value, here altogether comprising k voltage peak time point from m in action potential sequence M1To mk, define an action electricity
Bit sequence is N, wherein niThe time point of i-th of voltage peak is reached in expression action potential sequence N, here action potential sequence
Altogether comprising r voltage peak time point from n in N1To nr
M={ m1, m2, m3..., mk}
N={ n1, n2, n3..., nr}
2. defining a kernel function h (t)
Wherein it is time constant
3. defining two the series fs (t, M) and f (t, N) unrelated with the shape of action potential
4. it is d to define the distance between two groups of action potential sequences M and N
5. the calculating of above-mentioned distance d can be converted to following form
6. wherein mjThe time point of j-th of voltage peak is reached in expression action potential sequence M, here action potential sequence
Altogether comprising k voltage peak time point from m in M1To mk, wherein njIt indicates to reach j-th of Voltage Peak in action potential sequence N
The time point of value, here altogether comprising r voltage peak time point from n in action potential sequence N1To nr
7. a d has been mapped between [0,1], a fitness value score is obtained
The distance between two groups of action potential sequences d can thus be illustrated with the fitness value between 0-1.
Wherein in order to solve the problems, such as that calculating individual adaptation degree value speed is very slow in genetic algorithm, for each individual
It is required for HH equation of calculating to obtain its fitness value, is assigned in a GPU core using by each individual, to
The Calculation bottleneck caused when individual amount is very more in population is greatly overcome, big population can be easily reached at high speed
Requirement.
Beneficial effect：
Software uses paralleling genetic algorithm to solve neuron ion channel Parametric optimization problem, with prior art phase
Than, the invention has the advantages that：
(1) method for the pseudo- experimental data of generation that this algorithm provides can clearly and succinctly check algorithm effect；
(1) paralleling genetic algorithm is mutually tied the high-speed parallel of parallel computer with the natural concurrency of genetic algorithm
It closes, using by the method for HH equation parallel computations, substantially increases operational efficiency；
Description of the drawings
In order to more clearly explain the embodiment of the invention or the technical proposal in the existing technology, to embodiment or will show below
There is required attached drawing in technology description to be briefly described.
Fig. 1 is the structure chart of neuron parameter fitting device in embodiment of the present invention.
Fig. 2 is the broad flow diagram of genetic algorithm used in neuron parameter fitting device in embodiment of the present invention.
Fig. 3 is the broad flow diagram for generating pseudo- experimental data in embodiment of the present invention in neuron parameter fitting device.
Specific implementation mode
Following will be combined with the drawings in the embodiments of the present invention, and technical solution in the embodiment of the present invention carries out clear, complete
Site preparation describes.
Fig. 1 is the structure chart of neuron parameter fitting device of the embodiment of the present invention, and main includes inputting or obtaining data mould
Block 11, genetic algorithm module 12, parameter display module 13, wherein：
It is to import true biological experiment data or given birth to by this software to input or obtain 11 main function of data module
At pseudo- experimental data, data include spike time point datas and current data.
The generation of pseudo- experimental data is as shown in figure 3, comprise the following steps：
The input of each section of stimulated current of step S31.
Some parameters of step S32 selection ion channels and the corresponding conductance parameter value of input and HH equations.
Step S33 generates pseudo- experimental data and visually galvanic current and voltage.
The main function of genetic algorithm module 12 is the fitting that neuron parameter is carried out with genetic algorithm.
The operational process of genetic algorithm is as shown in Fig. 2, comprise the following steps：
Step S21 generates initial population, calculates fitness value.
Step S22 carries out selection operation according to fitness value, to the individual in population.
Step S23 carries out carry out crossover operation according to fitness value and crossover probability to the individual in population.
Step S24 carries out carry out mutation operation according to fitness value and mutation probability to the individual in population.
The main function of genetic algorithm module 13 be using genetic algorithm carry out neuron parameter fitting often for best
As a result, and by optimized parameter substitute into HH equation calculation voltage values, compared with pseudo- experimental data.With the first two conductance two-dimensional visualization
In addition the variation of individual also uses PCA by every group of conductance parameter dimensionality reduction to two-dimensional visualization, illustrate each individual in every generation
Scoring event and marked minimum value, in addition also show it is every operation an iteration needed for time.
Claims (5)
1. a kind of neuron parameter fitness method, which is characterized in that include the following steps：
Initial population is generated according to the parameter initialization population of input first, what each individual in population actually represented is exactly
One group of parameter also needs one group of true peak hour point data of input, this group of data to be measured and recorded by true biological experiment
It obtains or algorithm generates；Each individual in population generates one group of peak hour point data after substituting into HH equations, according to
This group of peak hour point data and the true spiking data for starting input calculate the corresponding fitness value of each individual；Then basis
Fitness value carries out genetic manipulation, final output optimum individual, i.e., optimal conductance parameter.
2. according to the method described in claim 1, it is characterized in that：When in population individual amount it is many and caused by calculate and adapt to
The process that each individual calculates fitness value is put into GPU cores and carries out by the bottleneck of angle value.
3. according to the method described in claim 1, it is characterized in that：Peak hour point data is generated by following methods：
The value of conductance parameter is first given, reselection needs the ion channel type being fitted, then substitutes into following HH equations and obtain
Data；
Wherein CMIndicate that capacitance, V refer to voltage value,It is electric conductivity value, m, n, h is the differential equation about t, and I refers to current
The stimulated current of neuronal cell is given, includes a sodium channel and a potassium channel in this equation, the m in above formula, n, h divide
The opening and closing dynamic process of these three ion channels is not reflected；αn, βnIndicate the rate constant of associated ion channel.
4. according to the method described in claim 1, it is characterized in that：
Parameter fitting is carried out using paralleling genetic algorithm, is as follows：
1) real coding is carried out to conductance parameter first, x=[x1, x2, x3, x4, x5] wherein x1~x5 represents five conductance ginsengs
Number, including a sodium channel, a calcium channel and three potassium channels, each group of conductance parameter represent an individual, later
Generate the initial population A for including many this individuals；
2) it selects, excellent individual is replicated according to fitness value, eliminates poor individual：
A Small Population A1 is selected from initial population A, and population A 1 is divided equally into two parts a1, a2, compares a1, the correspondence position of a2
The size of individual fitness value is set, if the fitness value of the individual in a2 on some position is better than a1, with the individual in a2
Replace the individual of corresponding position on a1；
3) mixing intersects：
Mixing is carried out according to crossover probability to intersect, i.e., to each of a2 individual c, selected in a1 at random to each of a2 individuals
Two individuals n1, n2 are selected, according to formula Ki=n1*weight+n2* (1-weight) executes intersection, and weight is between 0-1
Random value；
4) it makes a variation：
Normal distribution variation is carried out according to mutation probability to each of a2 individuals, i.e., to each of a2 individuals KiThere are five bases
Because of position, i.e., the parameter value of five kinds ion channels, Ki=[x1, x2, x3, x4, x5], then being directed to gene position x1, (first ion is logical
Road), x1=weight, wherein weight Normal Distributions set the mean value of its Normal Distribution as (x1_Low+ here
X1_Up)/2, x1_Low indicates that the lower bound of first ion channel parameter value of neuron, x1_Up indicate first ion of neuron
The upper bound of channel parameters value
5) population exchanges：
From some individuals being introduced into a1 in some individual substitution population As, some individuals are introduced from A, replace some in a1
Individual；
6) return of population A 1 is obtained into new population A：
7) 2) -6 are repeated), the maximum genetic algebra until reaching setting returns the result.
5. according to the method described in claim 1, it is characterized in that,
The calculating of fitness value uses following method：
A) it is M, wherein m to define an action potential sequence firstiIt indicates to reach i-th voltage peak in action potential sequence
Time point, here altogether comprising k voltage peak time point from m in action potential sequence M1To mk, define an action potential sequence
It is classified as N, wherein niThe time point that i-th of voltage peak is reached in expression action potential sequence N, here in action potential sequence N
Include r voltage peak time point altogether from n1To nr
M={ m1, m2, m3..., mk}
N={ n1, n2, n3..., nr}
B) a kernel function h (t) is defined
Wherein τ is time constant
C) two the series fs (t, M) and f (t, N) unrelated with the shape of action potential are defined
D) it is d to define the distance between two groups of action potential sequences M and N
E) calculating of above-mentioned distance d is converted to following form
Wherein mjThe time point of j-th of voltage peak is reached in expression action potential sequence M, here in action potential sequence M altogether
Including k voltage peak time point is from m1To mk, wherein njThe time of j-th of voltage peak is reached in expression action potential sequence N
Point, here altogether comprising r voltage peak time point from n in action potential sequence N1To nr
F) between d being mapped to [0,1], a fitness value score is obtained
The distance between two groups of action potential sequences d is indicated with the fitness value between 0-1.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201810517988.7A CN108764474A (en) | 2018-05-25 | 2018-05-25 | A kind of neuron parameter fitness method |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201810517988.7A CN108764474A (en) | 2018-05-25 | 2018-05-25 | A kind of neuron parameter fitness method |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN108764474A true CN108764474A (en) | 2018-11-06 |
Family
ID=64006134
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201810517988.7A Pending CN108764474A (en) | 2018-05-25 | 2018-05-25 | A kind of neuron parameter fitness method |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN108764474A (en) |
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109670585A (en) * | 2018-12-29 | 2019-04-23 | 中国人民解放军陆军工程大学 | Neuron bionic circuit and neuromorphic system |
| CN112182974A (en) * | 2020-10-09 | 2021-01-05 | 中国人民解放军陆军工程大学 | Neuron circuit evolution design method |
Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN1450493A (en) * | 2003-04-25 | 2003-10-22 | 北京工业大学 | Nerve network system for realizing genetic algorithm |
| CN103105246A (en) * | 2012-12-31 | 2013-05-15 | 北京京鹏环球科技股份有限公司 | Greenhouse environment forecasting feedback method of back propagation (BP) neural network based on improvement of genetic algorithm |
-
2018
- 2018-05-25 CN CN201810517988.7A patent/CN108764474A/en active Pending
Patent Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN1450493A (en) * | 2003-04-25 | 2003-10-22 | 北京工业大学 | Nerve network system for realizing genetic algorithm |
| CN103105246A (en) * | 2012-12-31 | 2013-05-15 | 北京京鹏环球科技股份有限公司 | Greenhouse environment forecasting feedback method of back propagation (BP) neural network based on improvement of genetic algorithm |
Non-Patent Citations (3)
| Title |
|---|
| A. L. HODGKIN ET AL: "A quantitative description of membrane current and its application to conduction and excitation in nerve", BULLETIN OF MATHEMATICAL BIOLOGY * |
| CONOR HOUGHTON ET AL: "On the efficient calculation of van Rossum distances", NETWORK COMPUTATION IN NEURAL SYSTEMS * |
| 沈佳敏 等: "神经元参数自动拟合", 中国科技论文在线 * |
Cited By (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109670585A (en) * | 2018-12-29 | 2019-04-23 | 中国人民解放军陆军工程大学 | Neuron bionic circuit and neuromorphic system |
| CN109670585B (en) * | 2018-12-29 | 2024-01-23 | 中国人民解放军陆军工程大学 | Neuron bionic circuit and neuromorphic system |
| CN112182974A (en) * | 2020-10-09 | 2021-01-05 | 中国人民解放军陆军工程大学 | Neuron circuit evolution design method |
| CN112182974B (en) * | 2020-10-09 | 2022-04-26 | 中国人民解放军陆军工程大学 | Neuron circuit evolution design method |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| JP7235344B2 (en) | A Quantum Computation Method for Expressway Traffic Flow Distribution Simulation Considering Destination Selection | |
| CN102768701B (en) | High-voltage switch cabinet insulator electric field optimization method based on quantum genetic algorithm | |
| CN109242149A (en) | A kind of student performance early warning method and system excavated based on educational data | |
| CA2436352A1 (en) | Process and system for developing a predictive model | |
| Zhou et al. | Parameters inversion of high central core rockfill dams based on a novel genetic algorithm | |
| CN109360604A (en) | A kind of oophoroma molecule parting forecasting system | |
| CN104798043A (en) | Data processing method and computer system | |
| CN109507527B (en) | Fault positioning method for power distribution network containing distributed power supply of SNP system based on synaptic plasticity | |
| Harte et al. | DynaMETE: a hybrid MaxEnt‐plus‐mechanism theory of dynamic macroecology | |
| CN108764474A (en) | A kind of neuron parameter fitness method | |
| CN109886560A (en) | Distribution network transform measure and rate of qualified voltage index relevance method for digging and device | |
| CN110009181A (en) | Distribution network transform measure and mistake load figureofmerit relevance method for digging and device | |
| CN106599610A (en) | Method and system for predicting association between long non-coding RNA and protein | |
| Ma et al. | Multigroup strategy for well control optimization | |
| CN110298506A (en) | A kind of urban construction horizontal forecast system | |
| CN109886316B (en) | Transformer state parameter combination prediction method based on cloud system similarity weight distribution | |
| Hafezi et al. | Synthesizing population for microsimulation-based integrated transport models using Atlantic Canada micro-data | |
| CN110533341A (en) | A kind of Livable City evaluation method based on BP neural network | |
| Ferreira et al. | Evolutionary strategy for simultaneous optimization of parameters, topology and reservoir weights in echo state networks | |
| CN109063837A (en) | Genetic algorithm information flow network property analysis method based on complex network structures entropy | |
| Hillis et al. | To tree the truth: Biological and numerical simulations of phylogeny | |
| Tchorzewski | Systemic Evolution Algorithms for the design of the development of the electricity market | |
| CN109886460A (en) | The prediction technique of tunnel subsidence time series based on adaboost | |
| CN109558633A (en) | The intelligent oil pumping machine parameter regulation means of network are generated based on intensified learning and confrontation | |
| Hu et al. | An intelligent test paper generation method based on genetic particle swarm optimization |
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 | ||
| WD01 | Invention patent application deemed withdrawn after publication |
Application publication date: 20181106 |
|
| WD01 | Invention patent application deemed withdrawn after publication |