CN111210353B - Intelligent triggering and informing method - Google Patents
Intelligent triggering and informing method Download PDFInfo
- Publication number
- CN111210353B CN111210353B CN202010017083.0A CN202010017083A CN111210353B CN 111210353 B CN111210353 B CN 111210353B CN 202010017083 A CN202010017083 A CN 202010017083A CN 111210353 B CN111210353 B CN 111210353B
- Authority
- CN
- China
- Prior art keywords
- data
- curve
- frequency domain
- judging
- fluctuation
- 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.)
- Active
Links
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q40/00—Finance; Insurance; Tax strategies; Processing of corporate or income taxes
- G06Q40/06—Asset management; Financial planning or analysis
-
- 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
-
- 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)
- Theoretical Computer Science (AREA)
- Biophysics (AREA)
- Life Sciences & Earth Sciences (AREA)
- Health & Medical Sciences (AREA)
- Business, Economics & Management (AREA)
- General Physics & Mathematics (AREA)
- Computational Linguistics (AREA)
- Finance (AREA)
- Software Systems (AREA)
- Mathematical Physics (AREA)
- General Engineering & Computer Science (AREA)
- Computing Systems (AREA)
- Molecular Biology (AREA)
- General Health & Medical Sciences (AREA)
- Evolutionary Computation (AREA)
- Artificial Intelligence (AREA)
- Data Mining & Analysis (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Bioinformatics & Computational Biology (AREA)
- Development Economics (AREA)
- Evolutionary Biology (AREA)
- Accounting & Taxation (AREA)
- Biomedical Technology (AREA)
- Game Theory and Decision Science (AREA)
- Genetics & Genomics (AREA)
- Physiology (AREA)
- Entrepreneurship & Innovation (AREA)
- Economics (AREA)
- Human Resources & Organizations (AREA)
- General Business, Economics & Management (AREA)
- Operations Research (AREA)
- Technology Law (AREA)
- Strategic Management (AREA)
- Marketing (AREA)
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
Abstract
The invention provides an intelligent triggering and informing method, which comprises the following steps: s1: obtaining original data; s2: cleaning original data, denoising, and obtaining a smooth preliminary signal oscillogram by utilizing wavelet analysis; s3: preliminarily extracting curve characteristics; s4: obtaining curve characteristics in depth; s5: optimizing the depth curve fluctuation feature set obtained in the step S4 by using historical data, performing combined analysis on different fluctuation features, and gradually increasing the prediction probability; s6: analyzing the obtained depth curve fluctuation feature set to separate a plurality of different curve features, S7: and setting a trigger condition and distributing signals to the users. The method can prepare for extracting the characteristics of the historical data, can quickly make judgment according to the characteristics factors and the market changes according to the characteristics factors by iterative optimization, and can give user suggestions or trades.
Description
Technical Field
The invention relates to the technical field of intelligent delivery and treatment, in particular to an intelligent triggering and informing method.
Background
In recent years, with the development of technologies, high and new technologies such as big data, artificial intelligence, cloud computing and mobile internet appear, the technologies are naturally and deeply integrated with financial energy, and the development of big data can search and mine from massive financial data more quickly and better, capture curve characteristics and the like and acquire financial change rules; the appearance of artificial intelligence can play a subversive role in the development of finance, and the artificial intelligence is combined with the finance. The scale and the quality of the data determine the intelligent degree of artificial intelligence to a certain degree, finance is an industry for generating mass data, the mass data are preliminarily cleaned, mass high-quality data are obtained, and a machine is trained through a deep learning algorithm. Therefore, the original data application and generation are predicted, and the deep learning and the prediction of the future data are based on the original data learning and processing of massive high-quality data, so that the massive high-quality data becomes the key of the method. Professor in the department of computer science of the university of kanamailong said that: "big data will play two roles in the future, one is to play many functions; secondly, show more chances, increase commercial value. Combining data with AI, it is optimal to generate a circular interaction. "
The intelligent investment is an upgraded product on the basis of traditional finance mass data generation and by combining the rapid big data and artificial intelligence developed in the year. The intelligent delivery and treatment is based on the traditional financial delivery and treatment, utilizes the big data technology and artificial intelligence technology which are developed at high speed in recent years, combines the data and the technology, and completes the financial and consultant service provided manually in the past through an algorithm and a product. The method can utilize massive data analysis to perform algorithm optimization and generate a quantitative model on the basis, and gives investment suggestions and transactions according to the needs of users, expected benefits and risks born by the users. And provides corresponding change signals according to the real-time change of market data, so that the user can manage the own wealth according to the optimal scheme.
Because the financial system is an open and complex chaotic system, fragment information explodes in the current era, investment opportunities vanish in the short term, and an artificial investment mode is difficult to effectively capture the investment opportunities. And the intelligent investment and customer setting-up a quantitative transaction strategy model by means of an investment theory, and then inputting variables such as risk preference of investors, financial conditions, financing planning and the like into the model to generate an automatic, intelligent and personalized asset configuration suggestion for the user. Domestic household wealth is steadily increased, the intermediate-quality stage is gradually enlarged, the wealth management market space is huge, but an investment channel is scarce, internet financing begins to be widely accepted and popular through a round of P2P market gift washing, meanwhile, the risk consciousness of the public is also improved, and the young generation is more approved to internet wealth management. Under this background, at the end of 2014, the concept of intelligent investment begins to be introduced into China, then a large number of scientific and technological enterprises begin to appear, and traditional financial institutions also lay out intelligent investment directions vigorously after the next half of 2015. Although the development of domestic intelligent investment follows the United states, the development degree of the financial market, tax system, supervision difference and other factors are different to a certain extent. From the perspective of participation in a subject and entering time, domestic intelligent commissioning and caring companies can be divided into three types, namely independent innovation companies, internet sponsor layouts and traditional financial institution layouts. If the investment target and the platform form are divided into four categories according to the user positioning, the investment target and the platform form, the categories comprise a 2C innovation platform, an asset allocation proposal platform, an active investment proposal platform and a comprehensive financing platform.
However, the intelligent delivery method in the prior art cannot rapidly learn and correct, the accuracy of market judgment is not too high, and certain hysteresis exists, so that the judgment of the user cannot be rapidly fed back to the user.
The method is used as an intelligent investment method, algorithm processing and processing are carried out on traditional market and financial derivative data, some model characteristics are obtained according to a certain algorithm and a quantitative model, and suggestions and signals are provided for users by combining the requirements of the users, the requirements of implementation and the development of technology.
Disclosure of Invention
In order to solve the technical problems in the prior art, the invention provides a method for intelligently triggering and notifying, in particular to a method for intelligently triggering and notifying the existing real-time data by forming specific curve characteristics based on historical data and matching the existing real-time data in real time, and the specific technical scheme is as follows:
in one aspect, the present invention provides an intelligent triggering and notification method, which is characterized by comprising the following steps: s2: cleaning original data, denoising, and obtaining a smooth preliminary signal oscillogram by utilizing wavelet analysis;
after the original data are input, the wrong original data are removed to obtain primary data; analyzing the frequency and time of the data according to the initial characteristics of the preliminary data, removing data outside a frequency and smooth curve, then removing noise points which are not on a transformation curve according to wavelet transformation to obtain pure data, and decomposing the waveform of the data into a related waveform frequency domain with simple waveform according to Fourier transformation;
s3: preliminarily extracting curve characteristics:
obtaining a group of frequency domain values by using the data of the preliminary signal oscillogram in the step S2, grouping according to the association probability of the comparison frequency domain values, obtaining discrete frequency domain values associated with the associated probability after classifying the frequency domain values, and taking the discrete frequency domain values as preliminary curve fluctuation characteristics;
s4: depth acquisition curve characteristics:
diffusing the preliminary curve fluctuation characteristics obtained in the step S3 to the whole historical data to obtain a set of depth curve fluctuation characteristics;
s5: optimizing the depth curve fluctuation feature set obtained in the step S4 by using the existing result, and performing combined analysis on different fluctuation features to gradually improve the prediction probability;
s6: analyzing the acquired depth curve fluctuation feature set by using a neural network algorithm, separating a plurality of different curve features which are respectively a first curve feature code and a second curve feature code … … Nth curve feature code, and judging superposition of features of real-time data according with one or more curves according to the first curve feature code and the second curve feature code … … Nth curve feature code;
s7: setting a trigger condition according to the existing data, and distributing a signal to the user if the real-time data in the step S6 conforms to the superposition of the characteristics of one or more curves.
Preferably, the original data with the error removed in step S2 is removed according to the timestamp of the original data, the upper and lower limits of the original data, the opening price, and the closing price.
Preferably, the specific step of obtaining a set of frequency domain values in step S3 is: and (4) extracting the waveform frequency domain value of the simple signal in the preliminary signal waveform diagram data obtained in the step (S2), performing discrete time domain Fourier transform on the data of the complex signal aperiodic waveform data, and separating two or more groups of waveform frequency domain values of the simple signal to obtain a group of frequency domain values.
Preferably, the depth-obtaining curve feature in step S4 includes the specific steps of: predicting the data that has occurred (predicting the data that has occurred, for example, my waveform map formed with data from 2011 to 2015, which can be used for verifying the data from 2016-2017) in the preliminary curve fluctuation characteristics obtained in step S3, in the future, and using the predicted values as neurons of the neural network, gradually increasing the number of neurons (the result obtained before is a set of waveforms, each successful waveform is an element of the set, each element is a neuron) before inputting the correct result and outputting the correct result (the data from 2011 to 2015 at the time of the previous data determination, for example, a waveform occurs, the probability of the next waveform occurrence is 50%, verifying the data from 2016 to 2017 with this conclusion, and if this probability is the correct result) and gradually increasing the success rate of the determination output, when the power reaches a certain threshold value, recording the numerical characteristics of the group of neurons to obtain a set of depth curve fluctuation characteristics;
preferably, the step of determining the prejudged value is to set an initial value, prejudge by using historical data, modify the initial value according to a prejudged result, and determine the modified initial value as the prejudged value when the prejudged success rate reaches more than 60%.
Preferably, the process of optimizing the set of fluctuation features of the obtained depth curve in step S5 specifically includes: setting the maximum iteration number M of the genetic iteration algorithm, judging input variables and output results according to a set of fluctuation characteristics of a depth curve and judging cross results, when the output results are larger than a certain threshold value, proposing the cross factors, judging the whole data flow by using the cross factors, gradually modifying fixed variables of the cross factors, gradually modifying the fixed variables of the cross factors to obtain the success rate of the results, and further carrying out deep crossing on the cross factors after N cross factors are obtained.
Compared with the prior art, the invention has the following beneficial effects:
(1) the method for forming the specific curve characteristics based on the historical data and intelligently triggering and informing the real-time matched existing real-time data improves the efficiency of cleaning the historical data, can prepare for extracting the characteristics of the historical data, can optimize the factors of the characteristics according to iteration, can quickly make judgment according to market changes according to the characteristic factors, and can quickly give user suggestions or transactions.
(2) The method provided by the invention can be used for fast learning and correction, has high accuracy in market judgment and no hysteresis, and can feed back the judgment of the user to the user fast.
(3) The method provided by the invention can improve the efficiency and accuracy of cleaning the historical data, can also accurately extract the curve characteristics of the historical data, can perform self-optimization through continuous iteration to improve the preparation rate of judgment, can also quickly judge the trend of the data according to real-time data, can quickly send messages to users, and provides reference and decision for the users.
Drawings
FIG. 1 is a block diagram of a method framework for intelligent triggering and notification provided by the present invention;
Detailed Description
The following describes in detail an embodiment of the present invention with reference to fig. 1.
Referring to fig. 1, fig. 1 is a block diagram of a method for intelligent triggering and notification provided by the present invention; in one aspect, the invention provides a method for forming specific curve characteristics based on historical data to match with the existing real-time data intelligent trigger and notification in real time, which comprises the following steps:
s1: obtaining original data;
s2: cleaning original data, denoising, and obtaining a smooth preliminary signal oscillogram by utilizing wavelet analysis;
after the original data are input, the wrong original data are removed to obtain primary data; analyzing the frequency and time of the data according to the initial characteristics of the preliminary data, removing data outside a frequency and smooth curve, then removing noise points which are not on a transformation curve according to wavelet transformation to obtain pure data, and decomposing the waveform of the data into a related waveform frequency domain with simple waveform according to Fourier transformation;
s3: preliminarily extracting curve characteristics:
obtaining a group of frequency domain values by using the data of the preliminary signal oscillogram in the step S2, grouping according to the association probability of the comparison frequency domain values, obtaining discrete frequency domain values associated with the associated probability after classifying the frequency domain values, and taking the discrete frequency domain values as preliminary curve fluctuation characteristics;
s4: depth acquisition curve characteristics:
diffusing the preliminary curve fluctuation characteristics obtained in the step S3 to the whole historical data to obtain a set of depth curve fluctuation characteristics;
s5: optimizing the depth curve fluctuation feature set obtained in the step S4 by using the existing result, and performing combined analysis on different fluctuation features to gradually improve the prediction probability;
s6: analyzing the acquired depth curve fluctuation feature set by using a neural network algorithm, separating a plurality of different curve features which are respectively a first curve feature code and a second curve feature code … … Nth curve feature code, and judging superposition of features of real-time data according with one or more curves according to the first curve feature code and the second curve feature code … … Nth curve feature code;
s7: setting a trigger condition according to the existing data, and distributing a signal to the user if the real-time data in the step S6 conforms to the superposition of the characteristics of one or more curves.
In a preferred embodiment, the original data with errors removed in step S2 is removed according to the time stamp of the original data and the conditions for determining the upper and lower limits, the opening price and the closing price of the original data.
As a preferred embodiment, the specific steps of obtaining a set of frequency domain values in step S3 provided by the present invention are: and (4) extracting the waveform frequency domain value of the simple signal in the preliminary signal waveform diagram data obtained in the step (S2), performing discrete time domain Fourier transform on the data of the complex signal aperiodic waveform data, and separating two or more groups of waveform frequency domain values of the simple signal to obtain a group of frequency domain values. The waveform frequency domain value in the invention is a group of waveform diagrams related to amplitude, frequency, coefficient and constant.
As a preferred embodiment, the depth acquisition of the curve feature in step S4 provided by the present invention specifically includes the following steps: and (4) performing future prejudgment on the generated data according to the initial data in the preliminary curve fluctuation characteristics obtained in the step (S3), using a prejudgment value as a neuron of the neural network, gradually increasing the number of the neurons and gradually improving the success rate of judgment output before a correct result is input and output, and recording the numerical characteristics of the group of neurons to obtain a depth curve fluctuation characteristic set when the power reaches a certain threshold value.
Specifically, data that has occurred (for example, i can verify data in 2016-2017 with a waveform diagram formed by data in 2011-2015) is predicted in the future according to the initial data in the preliminary curve fluctuation feature obtained in step S3, and the predicted value is used as a neuron of the neural network, the number of neurons is gradually increased (the previously obtained result is a set of waveforms, each successful waveform is an element of the set, each element is a neuron) before data is input into a correct result and the correct result is output (the data in 2011-2015 at the time of the previous data judgment, for example, data in one waveform occurs, the probability of the next waveform occurs is 50%, the data in 2016-2017 is verified with the result, and if the probability is the correct result), and the success rate of judgment output is gradually increased, when the power reaches a certain threshold value, recording the numerical characteristics of the group of neurons to obtain a set of depth curve fluctuation characteristics;
as a preferred embodiment, the pre-judgment value determining step provided by the present invention is to set an initial value, then pre-judge by using historical data, modify the initial value according to the pre-judgment result, and determine the value as the pre-judgment value when the pre-judgment success rate reaches above 60%.
As a preferred embodiment, the process for optimizing the set of fluctuation features of the obtained depth curve in step S5 provided by the present invention specifically includes: setting the maximum iteration number M of the genetic iteration algorithm, judging input variables and output results according to a set of fluctuation characteristics of a depth curve and judging cross results, when the output results are larger than a certain threshold value, proposing the cross factors, judging the whole data flow by using the cross factors, gradually modifying fixed variables of the cross factors, gradually modifying the fixed variables of the cross factors to obtain the success rate of the results, and further carrying out deep crossing on the cross factors after N cross factors are obtained.
Specifically, the intelligent triggering and notifying method of the invention comprises the following steps:
1. the method comprises the steps of firstly cleaning data, removing noise from the obtained historical data, filtering out data which accord with a waveform within a period of time by utilizing wavelet analysis, wherein noise outside the waveform can be understood as an accident or an emergency, and filtering out the noise first, so that a relatively smooth waveform diagram is obtained.
2. Fourier transform is carried out on the data with fixed period by the processed data to obtain a group of frequency domain values, proportional discrete frequency domain values are obtained after the frequency domain values are classified, and the frequency domain values can be understood as the fluctuation characteristics of the curves
3. The time period is enlarged to obtain curve characteristics in a longer range, iteration and optimization are carried out on the curve characteristics by using a genetic algorithm, similar curve characteristics and the same curve characteristics which occur historically are used for prediction, and meanwhile, the prediction probability is obtained by comparing the curve characteristics with the results. The obtained fluctuation characteristics are optimized by using the existing results, different fluctuation characteristics are combined and analyzed, the prediction probability is gradually improved, and meanwhile, a neural network algorithm is used for analyzing the generated curve to see whether the combination of a plurality of different curve characteristics can be separated or not. When the method is applied to a real-time curve, a single thread analyzes single curve characteristics by using multiple threads, a thread pool is used for adjusting according to the number of the curve characteristics which are possibly generated, and the further optimization is to accelerate an algorithm by using a GPU (graphics processing unit), so that a signal result is obtained at a millisecond level.
4. When a certain algebra of genetic algorithm and neural network algorithm iteration is utilized, for example, the judgment probability of the signal is set to be 70%, the factor is used as an available factor to be put into a factor library to monitor real-time data, and meanwhile, the historical factor is continuously subjected to iterative optimization to improve the probability of occurrence prediction. When the real-time data triggers the condition of the factor, if the user subscribes to the factor, the signal is immediately pushed to the user to remind the user to operate or care about what happens. Signals are sent from trigger to capture using multithreading and GPU acceleration to the millisecond level.
The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention, and various modifications and changes may be made by those skilled in the art. Any modification, equivalent replacement, or improvement made within the spirit and principle of the present invention shall fall within the protection scope of the present invention.
Claims (1)
1. A method for intelligent triggering and notification, characterized by comprising the following steps:
s1: obtaining original data;
s2: cleaning original data, denoising, and obtaining a smooth preliminary signal oscillogram by utilizing wavelet analysis;
after the original data are input, the wrong original data are removed to obtain primary data; analyzing the frequency and time of the data according to the initial characteristics of the preliminary data, removing data outside a frequency and smooth curve, then removing noise points which are not on a transformation curve according to wavelet transformation to obtain data without noise points, and decomposing the waveform of the data into a waveform frequency domain with simple waveform according to Fourier transformation;
s3: preliminarily extracting curve characteristics:
obtaining a group of frequency domain values by using the data of the preliminary signal oscillogram in the step S2, grouping according to the association probability of the comparison frequency domain values, obtaining discrete frequency domain values associated with the associated probability after classifying the frequency domain values, and taking the discrete frequency domain values as preliminary curve fluctuation characteristics;
s4: depth acquisition curve characteristics:
diffusing the preliminary curve fluctuation characteristics obtained in the step S3 to the whole historical data to obtain a set of depth curve fluctuation characteristics;
s5: optimizing the depth curve fluctuation feature set obtained in the step S4 by using the existing result, and performing combined analysis on different fluctuation features to gradually improve the prediction probability;
s6: analyzing the acquired depth curve fluctuation feature set by using a neural network algorithm, separating a plurality of different curve features which are respectively a first curve feature code and a second curve feature code … … Nth curve feature code, and judging superposition of features of real-time data according with one or more curves according to the first curve feature code and the second curve feature code … … Nth curve feature code;
s7: setting a trigger condition according to the existing data, and distributing a signal to the user if the real-time data in the step S6 conforms to the superposition of the characteristics of one or more curves;
the specific steps of obtaining a set of frequency domain values in step S3 are:
extracting the waveform frequency domain values of the simple signals in the preliminary signal waveform diagram data obtained in the step S2, performing discrete time domain Fourier transform on the aperiodic waveform data of the complex signals, and separating two or more groups of waveform frequency domain values of the simple signals to obtain a group of frequency domain values;
in step S4, obtaining the curve feature deeply includes the specific steps: performing future prejudgment on the generated data according to the initial data in the preliminary curve fluctuation features obtained in the step S3, using a prejudgment value as a neuron of a neural network, gradually increasing the number of the neurons and gradually improving the success rate of judgment output before a correct result is input and output, and recording the numerical value features of the neurons to obtain a depth curve fluctuation feature set when the power reaches a certain threshold value;
in the step S2, the original data with errors removed are removed according to the time stamp of the original data and the conditions for judging the upper and lower limits, the opening price and the closing price of the original data;
the pre-judging value determining step is that an initial value is set firstly, then the historical data is used for pre-judging, the initial value is modified according to the pre-judging result, and when the pre-judging success rate reaches more than 60%, the modified initial value is determined as the pre-judging value;
the process of optimizing the set of the obtained depth curve fluctuation features in step S5 specifically includes: setting the maximum iteration number M of a genetic iteration algorithm, judging input variables and output results according to a set of fluctuation characteristics of a depth curve and judging cross results, when the output results are larger than a certain threshold value, proposing the cross factors, judging the whole data flow by using the cross factors, gradually modifying fixed variables of the cross factors, gradually modifying the fixed variables of the cross factors, and further performing depth crossing on the cross factors after N cross factors are obtained;
when a certain algebra of genetic algorithm and neural network algorithm iteration is utilized, the judgment probability of the signal is set to be 70%, the factor is used as an available factor to be placed in a factor library to monitor real-time data, and meanwhile, the historical factor is continuously subjected to iteration optimization to improve the probability of occurrence prediction; when the real-time data triggers the condition of the factor, if the user subscribes to the factor, the signal is immediately pushed to the user to remind the user to operate or care about the occurrence; signals are sent from trigger to capture using multithreading and GPU acceleration to the millisecond level.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202010017083.0A CN111210353B (en) | 2020-01-08 | 2020-01-08 | Intelligent triggering and informing method |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202010017083.0A CN111210353B (en) | 2020-01-08 | 2020-01-08 | Intelligent triggering and informing method |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN111210353A CN111210353A (en) | 2020-05-29 |
| CN111210353B true CN111210353B (en) | 2021-02-26 |
Family
ID=70789005
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202010017083.0A Active CN111210353B (en) | 2020-01-08 | 2020-01-08 | Intelligent triggering and informing method |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN111210353B (en) |
Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN104268063A (en) * | 2014-07-07 | 2015-01-07 | 上海交通大学 | Trend analysis based performance prediction method and system of distributed system |
| CN109639450A (en) * | 2018-10-23 | 2019-04-16 | 平安壹钱包电子商务有限公司 | Fault alarming method, computer equipment and storage medium neural network based |
| CN110288482A (en) * | 2019-07-02 | 2019-09-27 | 欧冶云商股份有限公司 | Steel mill's futures exchange method and system |
Family Cites Families (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US10453142B2 (en) * | 2009-02-11 | 2019-10-22 | Johnathan Mun | System and method for modeling and quantifying regulatory capital, key risk indicators, probability of default, exposure at default, loss given default, liquidity ratios, and value at risk, within the areas of asset liability management, credit risk, market risk, operational risk, and liquidity risk for banks |
| US20140222716A1 (en) * | 2013-02-01 | 2014-08-07 | Michael A. Joplin | Methods And Systems For Processing Debt Portfolios |
| CN105844347A (en) * | 2016-03-16 | 2016-08-10 | 中山大学 | Individual stock index prediction method based on Lagrange multiplier method |
| CN108346099A (en) * | 2018-02-08 | 2018-07-31 | 上海宽全智能科技有限公司 | Hedging transaction analysis method, equipment and storage medium based on wavelet analysis |
| CN108596765A (en) * | 2018-04-28 | 2018-09-28 | 国信优易数据有限公司 | A kind of Electronic Finance resource recommendation method and device |
| CN108711107A (en) * | 2018-05-25 | 2018-10-26 | 上海钱智金融信息服务有限公司 | Intelligent financing services recommend method and its system |
| CN108921701A (en) * | 2018-07-19 | 2018-11-30 | 安徽磐众信息科技有限公司 | A kind of stock index futures price expectation method based on EDM algorithm and BP neural network |
| CN109190937A (en) * | 2018-08-16 | 2019-01-11 | 深圳前海乘方互联网金融服务有限公司 | A kind of investment value assessment system |
-
2020
- 2020-01-08 CN CN202010017083.0A patent/CN111210353B/en active Active
Patent Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN104268063A (en) * | 2014-07-07 | 2015-01-07 | 上海交通大学 | Trend analysis based performance prediction method and system of distributed system |
| CN109639450A (en) * | 2018-10-23 | 2019-04-16 | 平安壹钱包电子商务有限公司 | Fault alarming method, computer equipment and storage medium neural network based |
| CN110288482A (en) * | 2019-07-02 | 2019-09-27 | 欧冶云商股份有限公司 | Steel mill's futures exchange method and system |
Also Published As
| Publication number | Publication date |
|---|---|
| CN111210353A (en) | 2020-05-29 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Li et al. | Heterogeneous ensemble for default prediction of peer-to-peer lending in China | |
| Harvey et al. | Automated feature design for numeric sequence classification by genetic programming | |
| Nadkarni et al. | Combining NeuroEvolution and Principal Component Analysis to trade in the financial markets | |
| CN106600369A (en) | Real-time recommendation system and method of financial products of banks based on Naive Bayesian classification | |
| CN110998608B (en) | Machine learning system for various computer applications | |
| CN110084610A (en) | A kind of network trading fraud detection system based on twin neural network | |
| Arinze | Selecting appropriate forecasting models using rule induction | |
| Foroghi et al. | Applying decision tree to predict bankruptcy | |
| Guo | Oil price forecast using deep learning and ARIMA | |
| CN109102386A (en) | Risk control method and device based on big data | |
| CN113344700A (en) | Wind control model construction method and device based on multi-objective optimization and electronic equipment | |
| Faraz et al. | Stock market prediction-by-prediction based on autoencoder long short-term memory networks | |
| Cheong et al. | Interpretable stock anomaly detection based on spatio-temporal relation networks with genetic algorithm | |
| Lee et al. | Hidden markov models for forex trends prediction | |
| Huang et al. | A learning-based contrarian trading strategy via a dual-classifier model | |
| CN111210353B (en) | Intelligent triggering and informing method | |
| CN115760127A (en) | Transaction fraud detection method and system based on rule attention mechanism | |
| Rubinstein et al. | Time series forecasting of crude oil consumption using neuro-fuzzy inference | |
| CN112132679A (en) | Product recommendation method and device | |
| Pan | Design of financial management model using the forward neural network based on particle swarm optimization algorithm | |
| Zhai et al. | Big data analysis of accounting forecasting based on machine learning | |
| CN113781056A (en) | Method and device for predicting user fraud behavior | |
| CN114529063A (en) | Financial field data prediction method, device and medium based on machine learning | |
| Li et al. | Evaluating forecasting algorithm of realistic datasets based on machine learning | |
| Janosek et al. | Knowledge discovery in dynamic data using neural networks |
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 | ||
| TR01 | Transfer of patent right |
Effective date of registration: 20211220 Address after: 518000 unit B, 46th floor, China energy storage building, 3099 Keyuan South Road, high tech community, Yuehai street, Nanshan District, Shenzhen, Guangdong Patentee after: Gaoying International Innovation Technology (Shenzhen) Co.,Ltd. Address before: No.1-14, 22 / F, block a, building 9, shenzhenwan science and technology ecological park, Yuehai street, Nanshan District, Shenzhen, Guangdong 518000 Patentee before: GAOYING QUANTITATIVE CLOUD TECHNOLOGY (SHENZHEN) CO.,LTD. |
|
| TR01 | Transfer of patent right |