Abnormal prediction method and device for R&D progress, storage medium and electronic device

Abstract

The present application discloses a method and device for predicting abnormalities in R&D progress, a storage medium and an electronic device, and relates to the field of artificial intelligence. The method comprises: obtaining a first target data set, wherein the first target data set at least includes a statistical information value of the target data, wherein the target data is used to represent the fluctuation characteristics of the R&D progress data, and the statistical information value is used to represent the attribute information of the target data; processing the data in the first target data set according to a preset algorithm one and an automatic encoder to obtain a second target data set, wherein the fluctuation characteristics of the data in the second target data set are stronger than the fluctuation characteristics of the data in the first target data set; predicting abnormalities in the R&D progress according to the preset algorithm two and the second target data set. Through the present application, the problem that it is difficult to make abnormal warnings in advance when predicting abnormalities in the R&D progress by setting thresholds in the related art is solved, resulting in low R&D efficiency.

Description

Technical Field

[0001] The present application relates to the field of artificial intelligence, and more specifically, to a method and device for predicting abnormalities in research and development progress, a storage medium, and an electronic device. Background Art

[0002] R&D progress is an important data basis for refined management, and the R&D progress is composed of design progress, coding progress, and testing progress. In the current related technologies, abnormal detection of R&D progress is generally completed by setting thresholds to warn, that is, an alarm is generated based on the fact that the tasks of relevant staff have exceeded the prescribed period. Specifically, an R&D cycle model is formulated according to DoD (completion standard), and a detailed and complete R&D cycle will be obtained according to the R&D cycle model, that is, different stages need to be completed at fixed time nodes. Based on the R&D cycle model, the corresponding work cycle and days are pre-set for the R&D progress. If it exceeds the pre-set work cycle and days, it is determined that the R&D progress is lagging behind, and an alarm is issued.

[0003] However, there are obvious drawbacks in using the threshold-setting method to issue early warnings for R&D progress. Due to differences in R&D content, scale, resource investment, etc., the model cycles are different. Therefore, in the process of project advancement, potential progress anomalies may sometimes occur. If progress anomalies are judged only based on artificially set thresholds, when an abnormality warning is issued, the task progress has exceeded the specified time. Although the warning time can be advanced, this method has low flexibility and slow response speed. It can only set a fixed warning time according to the time given by the R&D cycle model, and cannot be distinguished according to the progress of different tasks.

[0004] When the relevant technology uses the method of setting thresholds to predict abnormalities in R&D progress, it is difficult to make abnormal warnings in advance, resulting in low R&D efficiency. No effective solution has been proposed yet. Summary of the invention

[0005] The main purpose of this application is to provide a method and device for predicting abnormalities in R&D progress, a storage medium and an electronic device, so as to solve the problem in the related art that when abnormalities in R&D progress are predicted by setting thresholds, it is difficult to make abnormal warnings in advance, resulting in low R&D efficiency.

[0006] In order to achieve the above-mentioned purpose, according to one aspect of the present application, a method for predicting abnormalities in R&D progress is provided. The method comprises: obtaining a first target data set, wherein the first target data set at least includes statistical information values ​​of target data, wherein the target data is used to represent the fluctuation characteristics of R&D progress data, and the statistical information values ​​are used to represent the attribute information of the target data; processing the data in the first target data set according to a preset algorithm one and an automatic encoder to obtain a second target data set, wherein the fluctuation characteristics of the data in the second target data set are stronger than the fluctuation characteristics of the data in the first target data set, and the automatic encoder is used to strengthen the fluctuation characteristics of the data in the first target data set when encoding the data in the first target data set; predicting abnormalities in the R&D progress according to a preset algorithm two and the second target data set.

[0007] Furthermore, before obtaining the first target data set, the method also includes: obtaining a third target data set, wherein the third target data set at least includes the target data; calculating statistical information values ​​of the target data based on the third target data set and a sliding window algorithm; and combining the statistical information values ​​of the target data to obtain the first target data set.

[0008] Furthermore, before obtaining the third target data set, the method also includes: obtaining original data in the original data set, wherein the original data is data representing the research and development progress; calculating the L2 norm and the norm change rate of the original data based on the original data and the sliding window algorithm; and combining the L2 norm and the norm change rate to obtain the third target data set.

[0009] Furthermore, before obtaining the original data in the original data set, the method also includes: calculating the amount of work completed in the R&D progress according to the target device; calculating the R&D progress according to the completed work and a preset formula; and converting the R&D progress into a vector representation to obtain the original data.

[0010] Furthermore, performing an abnormal prediction on the R&D progress according to preset algorithm two and the second target data set includes: determining an abnormal value in the second target data set according to the preset algorithm two; and performing an abnormal prediction on the R&D progress based on the abnormal value.

[0011] Furthermore, making an abnormal prediction of the R&D progress based on the abnormal value includes: if the difference between the abnormal value and the preset value is greater than a first preset value, determining that the possibility of an abnormality in the R&D progress is greater than a first preset probability; if the difference between the abnormal value and the preset value is less than a second preset value, determining that the possibility of an abnormality in the R&D progress is less than a second preset probability.

[0012] In order to achieve the above-mentioned purpose, according to another aspect of the present application, a device for predicting abnormalities in R&D progress is provided. The device comprises: a first acquisition unit, used to acquire a first target data set, wherein the first target data set at least includes statistical information values ​​of target data, wherein the target data is used to represent the fluctuation characteristics of R&D progress data, and the statistical information values ​​are used to represent the attribute information of the target data; a first processing unit, used to process the data in the first target data set according to a preset algorithm one and an automatic encoder to obtain a second target data set, wherein the fluctuation characteristics of the data in the second target data set are stronger than the fluctuation characteristics of the data in the first target data set, and the automatic encoder is used to strengthen the fluctuation characteristics of the data in the first target data set when encoding the data in the first target data set; a first prediction unit, used to predict abnormalities in the R&D progress according to a preset algorithm two and the second target data set.

[0013] Furthermore, the device also includes: a second acquisition unit, used to acquire a third target data set before acquiring the first target data set, wherein the third target data set at least includes the target data; a first calculation unit, used to calculate the statistical information value of the target data based on the third target data set and a sliding window algorithm; and a first combination unit, used to combine the statistical information values ​​of the target data to obtain the first target data set.

[0014] Furthermore, the device also includes: a third acquisition unit, used to acquire the original data in the original data set before acquiring the third target data set, wherein the original data is data representing the research and development progress; a second calculation unit, used to calculate the L2 norm and the norm change rate of the original data based on the original data and the sliding window algorithm; a second combination unit, used to combine the L2 norm and the norm change rate to obtain the third target data set.

[0015] Furthermore, the device also includes: a third calculation unit, used to calculate the amount of work completed in the R&D progress according to the target device before obtaining the original data in the original data set; a fourth calculation unit, used to calculate the R&D progress according to the completed work and a preset formula; a first conversion unit, used to convert the R&D progress into a vector representation to obtain the original data.

[0016] Furthermore, the first prediction unit includes: a first determination module, used to determine the abnormal value in the second target data set according to the preset algorithm 2; and a first prediction module, used to make an abnormal prediction of the research and development progress according to the abnormal value.

[0017] Furthermore, the first prediction module includes: a first processing submodule, used to determine that the possibility of an abnormality in the research and development progress is greater than a first preset probability if the difference between the abnormal value and the preset value is greater than the first preset value; and a second processing submodule, used to determine that the possibility of an abnormality in the research and development progress is less than a second preset probability if the difference between the abnormal value and the preset value is less than a second preset value.

[0018] In order to achieve the above-mentioned purpose, according to another aspect of the present application, a computer-readable storage medium is provided, wherein the storage medium includes a stored program, wherein the program executes any one of the above-mentioned methods for predicting abnormalities in R&D progress.

[0019] In order to achieve the above-mentioned purpose, according to another aspect of the present application, an electronic device is provided, which includes one or more processors and a memory, and the memory is used to store one or more programs, wherein when the one or more programs are executed by the one or more processors, the one or more processors implement any one of the above-mentioned methods for predicting abnormalities in R&D progress.

[0020] Through this application, the following steps are adopted: obtaining a first target data set, wherein the first target data set at least includes the statistical information value of the target data, wherein the target data is used to represent the fluctuation characteristics of the R&D progress data, and the statistical information value is used to represent the attribute information of the target data; processing the data in the first target data set according to the preset algorithm one and the automatic encoder to obtain a second target data set, wherein the fluctuation characteristics of the data in the second target data set are stronger than the fluctuation characteristics of the data in the first target data set, and the automatic encoder is used to strengthen the fluctuation characteristics of the data in the first target data set when encoding the data in the first target data set; performing abnormal prediction on the R&D progress according to the preset algorithm two and the second target data set, solving the problem that it is difficult to make abnormal warning in advance when the abnormal prediction of the R&D progress is made by setting a threshold in the related technology, resulting in low R&D efficiency. By processing the data in the first target data set obtained to obtain the second target data set, and performing abnormal prediction on the R&D progress according to the preset algorithm two and the second target data set, it is possible to make early warning of the abnormalities existing in the R&D progress, thereby achieving the effect of improving R&D efficiency. BRIEF DESCRIPTION OF THE DRAWINGS

[0021] The drawings constituting a part of the present application are used to provide a further understanding of the present application. The illustrative embodiments and descriptions of the present application are used to explain the present application and do not constitute an improper limitation on the present application. In the drawings:

[0022] Figure 1 is a flow chart of a method for predicting abnormality of R&D progress according to an embodiment of the present application;

[0023] Figure 2 is a schematic diagram of calculating the statistical information value of fluctuation data in an embodiment of the present application;

[0024] Figure 3 Schematic diagram of calculating data norm and norm change rate in an embodiment of the present application;

[0025] Figure 4 is a schematic diagram of detecting abnormal values ​​using an isolation forest algorithm in an embodiment of the present application;

[0026] Figure 5 is a schematic diagram of an optional abnormal prediction method for R&D progress provided according to an embodiment of the present application;

[0027] Figure 6 is a schematic diagram of an abnormality prediction device for R&D progress provided according to an embodiment of the present application;

[0028] Figure 7 is a schematic diagram of an electronic device provided according to an embodiment of the present application. DETAILED DESCRIPTION

[0029] It should be noted that, in the absence of conflict, the embodiments and features in the embodiments of the present application can be combined with each other. The present application will be described in detail below with reference to the accompanying drawings and in combination with the embodiments.

[0030] In order to enable those skilled in the art to better understand the solution of the present application, the technical solution in the embodiments of the present application will be clearly and completely described below in conjunction with the drawings in the embodiments of the present application. Obviously, the described embodiments are only part of the embodiments of the present application, not all of the embodiments. Based on the embodiments in the present application, all other embodiments obtained by ordinary technicians in this field without creative work should fall within the scope of protection of the present application.

[0031] It should be noted that the terms "first", "second", etc. in the specification and claims of the present application and the above-mentioned drawings are used to distinguish similar objects, and are not necessarily used to describe a specific order or sequential order. It should be understood that the data used in this way can be interchanged where appropriate, so that the embodiments of the present application described here. In addition, the terms "including" and "having" and any of their variations are intended to cover non-exclusive inclusions, for example, a process, method, system, product or device comprising a series of steps or units is not necessarily limited to those steps or units clearly listed, but may include other steps or units that are not clearly listed or inherent to these processes, methods, products or devices.

[0032] It should be noted that the user information (including but not limited to user device information, user personal information, etc.) and data (including but not limited to data for display, data for analysis, etc.) involved in this disclosure are all information and data authorized by the user or fully authorized by all parties.

[0033] The present invention is described below in conjunction with preferred implementation steps. Figure 1 is a flow chart of a method for predicting abnormality of R&D progress according to an embodiment of the present application, such as Figure 1 As shown, the method comprises the following steps:

[0034] Step S101, obtaining a first target data set, wherein the first target data set at least includes statistical information values ​​of target data, wherein the target data is used to represent the fluctuation characteristics of the R&D progress data, and the statistical information values ​​are used to represent the attribute information of the target data.

[0035] For example, the first target data set may be a fluctuation characteristic data set with statistical information, the target data may be R&D progress data with fluctuation characteristics, and the statistical information value may be the mean, standard deviation, minimum value, maximum value, quartiles and range of the R&D progress data with fluctuation characteristics. Therefore, it is necessary to obtain a fluctuation characteristic data set with statistical information.

[0036] Step S102, according to preset algorithm 1 and automatic encoder, the data in the first target data set is processed to obtain a second target data set, wherein the fluctuation characteristics of the data in the second target data set are stronger than the fluctuation characteristics of the data in the first target data set, and the automatic encoder is used to strengthen the fluctuation characteristics of the data in the first target data set when encoding the data in the first target data set.

[0037] For example, the preset algorithm 1 mentioned above can be an artificial neural network algorithm, specifically a long short-term memory network algorithm (LSTM) or a gated recurrent unit structure algorithm (GRU) or a recurrent neural network algorithm (RNN), and the above-mentioned autoencoder can be an encoder for encoding data in a fluctuation feature data set with statistical information. Although the progress data of the previous time window has a great influence on the progress data of the next time window, the influence between the R&D progress data separated by multiple time windows will gradually weaken, so the "forget gate" structure introduced in LSTM can be used to learn this feature. In addition, since the input vector of the autoencoder is the same as the target vector, the fluctuation information will also be highlighted in the process of encoding and decoding. Therefore, the fluctuation feature data set with statistical information is connected to the LSTM-based autoencoder, and after the learning and self-encoding of the "input gate" and "output gate" of LSTM, a data set with enhanced features is obtained, and there will be obvious time series features between the feature data of the R&D progress in the data set with enhanced features. Therefore, in the R&D progress, if the potential risk of progress lag is found, that is, after the abnormality is found in the progress, it can be highlighted in this step of the model learning process.

[0038] Step S103, performing abnormal prediction on the R&D progress according to the preset algorithm 2 and the second target data set.

[0039] For example, the above-mentioned preset algorithm 2 may be an isolation forest algorithm or a random forest algorithm. After the algorithm learns the features, the isolation forest algorithm or the random forest algorithm is used to detect whether there is an abnormality in the R&D progress.

[0040] Through the above steps S101 to S103, the second target data set is obtained by processing the data in the acquired first target data set, and the abnormality prediction of the R&D progress is performed according to the preset algorithm 2 and the second target data set, so that the abnormalities in the R&D progress can be warned in advance, thereby achieving the effect of improving the R&D efficiency.

[0041] Optionally, in the abnormality prediction method for R&D progress provided in an embodiment of the present application, before obtaining the first target data set, the method also includes: obtaining a third target data set, wherein the third target data set at least includes the target data; calculating the statistical information value of the target data based on the third target data set and the sliding window algorithm; and combining the statistical information values ​​of the target data to obtain the first target data set.

[0042] For example, the third target data set may be a fluctuation characteristic data set, which is composed of R&D progress data with fluctuation characteristics. After obtaining the fluctuation characteristic data set, the statistical information value of the fluctuation data in the time window is calculated according to the fluctuation characteristic data set and the sliding window algorithm. The calculation process is as follows: Figure 2As shown in the figure, N1 can represent the norm, and DN1 can represent the norm change rate. Among them, the statistical information values ​​include mean, standard deviation, minimum value, maximum value, quartile and range, and then the fluctuation characteristic data with statistical information are combined to obtain the fluctuation characteristic data set with statistical information.

[0043] Through the above scheme, according to the fluctuation characteristics of progress data and the sliding window algorithm, the information expression of the data set with fluctuation characteristics can be enriched.

[0044] Optionally, in the abnormal prediction method for R&D progress provided in the embodiment of the present application, before obtaining the third target data set, the method further includes: obtaining original data in the original data set, wherein the original data is data representing the R&D progress; and calculating the L2 norm and norm change rate of the original data based on the original data and the sliding window algorithm, the calculation process is as follows: Figure 3 As shown, N1 in the figure can represent the norm, DN1 can represent the norm change rate, T1 can represent the design progress, T2 can represent the coding progress, T3 can represent the test progress, and t1 can represent the data representing the design progress calculated at different times; the L2 norm and the norm change rate are combined to obtain the third target data set.

[0045] For example, the original data of R&D progress is obtained, and the L2 norm and norm change rate of the R&D progress data are calculated based on the original data of R&D progress, the sliding window algorithm and the L2 norm formula, and then the L2 norm and the norm change rate are combined into a data set with fluctuation characteristics. Among them, the norm is a function that assigns non-zero positive length or size to all vectors in the vector space. All norms are often used to measure the length or size of each vector in a vector space (or matrix), that is, the norm of a vector can be simply understood as the length of the vector, or the distance from the vector to the zero point, or the distance between the corresponding two points. The L2 norm is the 1 / 2 power of the sum of the squares of each element of the vector.

[0046] Through the above scheme, according to the sliding window algorithm and the L2 norm formula, the fluctuation characteristics of the data can be extracted from the original data representing the R&D progress.

[0047] Optionally, in the abnormal prediction method of R&D progress provided in the embodiment of the present application, before obtaining the original data in the original data set, the method also includes: calculating the amount of work completed in the R&D progress according to the target device; calculating the R&D progress according to the completed workload and a preset formula; converting the R&D progress into a vector representation to obtain the original data.

[0048] For example, before obtaining the original data of the R&D progress, it is necessary to prepare the R&D progress data. For example, the calculation method of the design progress can be the following steps: first, calculate the completed design workload based on the JIRA (project and transaction tracking tool) workload, and then calculate the current design progress based on the completed design workload. Among them, the specific calculation formula is: the design workload completed by the requirement sub-item = (JIRA workload completed by the design / total JIRA workload of the design) * the development workload of the requirement sub-item * the split coefficient; the design progress of the requirement item D1 = ∑the completed design workload of the requirement sub-item / the development workload of the requirement item * 100%. In addition, the calculation method of the coding progress and the test progress is the same as the calculation method of the design progress.

[0049] In summary, by calculating the workload, we can accurately obtain data representing the R&D progress.

[0050] Optionally, in the abnormal prediction method of R&D progress provided in the embodiment of the present application, abnormal prediction of R&D progress is performed according to preset algorithm two and the second target data set, including: determining abnormal values ​​in the second target data set according to preset algorithm two; and performing abnormal prediction of R&D progress based on the abnormal values.

[0051] For example, using the isolation forest algorithm or the random forest algorithm, outliers in a data set with enhanced features are detected. The specific process is as follows Figure 4 As shown, based on the abnormal value, detect whether there is an abnormality in the R&D progress.

[0052] Through the above scheme, if data features with potential risks are found in the R&D progress process, warnings can be issued in advance, that is, the risks or abnormal situations in the R&D progress can be advanced, so that the staff can sort out the progress in time and check for omissions and fill in the gaps, thereby ensuring the normal progress of the R&D process and the normal advancement of the project.

[0053] Optionally, in the abnormality prediction method for R&D progress provided in the embodiment of the present application, abnormality prediction of R&D progress based on abnormal value includes: if the difference between the abnormal value and the preset value is greater than the first preset value, then determining that the possibility of abnormality in the R&D progress is greater than the first preset probability; if the difference between the abnormal value and the preset value is less than the second preset value, then determining that the possibility of abnormality in the R&D progress is less than the second preset probability.

[0054] For example, if the difference between the detected abnormal value and 1 is larger, it means that the possibility of abnormality in the R&D progress is greater; if the difference between the detected abnormal value and 1 is smaller, it means that the possibility of abnormality in the R&D progress is smaller.

[0055] Through the above scheme, the possibility of abnormality in the R&D progress can be accurately determined.

[0056] Schematic diagram of the optional abnormal prediction method of R&D progress, such as Figure 5 As shown. For example, the R&D progress data is divided into design progress data D1, coding progress data D2, and test progress data D3 to express the progress status of each stage of a complete R&D cycle, that is, D1, D2, and D3 mark the progress data sets of three stages. Moreover, the data can be refined to each development team in units of days, which is convenient for fine management of development and testing personnel and accurate progress tracking. In addition, considering that the R&D progress data is a typical multidimensional time series, and in order to achieve the purpose of mining potential R&D progress risks, the model calculates the change value on the data time scale, extracts the statistical features in the change value, reconstructs the original data set, and captures the fluctuation characteristics of the data. With the help of deep neural networks, the feature expression of the R&D progress data is strengthened to form a feature data vector, and the fluctuation characteristics are weighted. Considering the multidimensional characteristics of the data, the isolation forest algorithm is used for anomaly detection to mine abnormal nodes in the data change process. Therefore, building an anomaly detection model is divided into the following three steps: feature extraction, data reconstruction, and anomaly detection.

[0057] Since the abnormal situation in the R&D progress is often a period of data where the changes between data items are different from the normal fluctuations, the data of the R&D progress is one of the most representative features of the abnormal data. First, the R&D progress data D1, coding progress data D2, and test progress data D3 are converted into vector representations, for example:

[0058] D1=[0.124,0.212,0342,0.453,0.458,...,0.877,0.996];

[0059] D2=[0.153,0.278,0.398,0.573,0.611,...0.853,0.998];

[0060] D3=[0.142,0.284,0.378,0.498,0.623,...0.889,0.999].

[0061] Then, the sliding window 1 is used to calculate the L2 norm and norm change of the progress data, and the fluctuation characteristics of the data are extracted. Assuming that the values ​​of the two time windows are both 2, each set of data is converted from 1*N dimensions to 2*N, and the R&D progress after extracting the data fluctuation characteristics is as follows:

[0062]

[0063]

[0064]

[0065] After extracting the fluctuation characteristics, the statistical information values ​​of the data in the time window are calculated, including the mean, standard deviation, minimum, maximum, quartile and range, and the data is converted from 2*N to 2*((N-1)*8) dimensions. In addition, since the design progress, coding progress and test progress are three groups of related but relatively independent data, considering the correlation between them, D1, D2 and D3 are concatenated to form a 6*((N-1)*8)-dimensional input data set. The sample data is shown below:

[0066]

[0067]

[0068]

[0069] Then, the fluctuation feature data set containing statistical information is connected to the LSTM-based autoencoder, and the feature vector of the fluctuation feature data is weighted to strengthen the fluctuation feature. Finally, the isolation forest is used to detect anomalies in the fluctuation feature data set. Therefore, by combining the artificial neural network algorithm and the isolation forest algorithm to detect anomalies in the R&D progress, the R&D progress data can be well utilized, the characteristics of the R&D progress data can be learned, and the information in the R&D progress data can be mined. The fluctuation of the R&D progress can be mainly analyzed to find the potential risks of the current progress. In addition, the R&D progress data can be analyzed more flexibly, and the fluctuation characteristics of the progress data in different periods of the R&D process can be different, so as to flexibly give early warnings.

[0070] In summary, the abnormal prediction method of the R&D progress provided by the embodiment of the present application obtains a first target data set, wherein the first target data set at least includes the statistical information value of the target data, wherein the target data is used to represent the fluctuation characteristics of the R&D progress data, and the statistical information value is used to represent the attribute information of the target data; according to the preset algorithm one and the automatic encoder, the data in the first target data set is processed to obtain the second target data set, wherein the fluctuation characteristics of the data in the second target data set are stronger than the fluctuation characteristics of the data in the first target data set, and the automatic encoder is used to strengthen the fluctuation characteristics of the data in the first target data set when encoding the data in the first target data set; according to the preset algorithm two and the second target data set, the abnormal prediction of the R&D progress is performed, which solves the problem that it is difficult to make an abnormal warning in advance when the abnormal prediction of the R&D progress is performed by setting a threshold in the related technology, resulting in low R&D efficiency. By processing the data in the acquired first target data set to obtain the second target data set, and according to the preset algorithm two and the second target data set, the abnormal prediction of the R&D progress is performed, so that the abnormalities existing in the R&D progress can be warned in advance, thereby achieving the effect of improving the R&D efficiency.

[0071] It should be noted that the steps shown in the flowcharts of the accompanying drawings can be executed in a computer system such as a set of computer executable instructions, and that, although a logical order is shown in the flowcharts, in some cases, the steps shown or described can be executed in an order different from that shown here.

[0072] The embodiment of the present application also provides an abnormality prediction device for R&D progress. It should be noted that the abnormality prediction device for R&D progress of the embodiment of the present application can be used to execute the abnormality prediction method for R&D progress provided by the embodiment of the present application. The abnormality prediction device for R&D progress provided by the embodiment of the present application is introduced below.

[0073] Figure 6 Schematic diagram of an abnormality prediction device for the development progress according to an embodiment of the present application. Figure 6 As shown, the device includes: a first acquisition unit 601, a first processing unit 602 and a first prediction unit 603.

[0074] Specifically, the first acquisition unit 601 is used to acquire a first target data set, wherein the first target data set at least includes a statistical information value of the target data, wherein the target data is used to represent the fluctuation characteristics of the R&D progress data, and the statistical information value is used to represent the attribute information of the target data;

[0075] A first processing unit 602 is used to process the data in the first target data set according to a preset algorithm 1 and an autoencoder to obtain a second target data set, wherein the fluctuation characteristics of the data in the second target data set are stronger than the fluctuation characteristics of the data in the first target data set, and the autoencoder is used to strengthen the fluctuation characteristics of the data in the first target data set when encoding the data in the first target data set;

[0076] The first prediction unit 603 is used to make an abnormal prediction on the R&D progress according to the preset algorithm 2 and the second target data set.

[0077] In summary, the apparatus for predicting anomalies in R&D progress provided by the embodiment of the present application obtains a first target data set through a first acquisition unit 601, wherein the first target data set at least includes a statistical information value of the target data, wherein the target data is used to represent the fluctuation characteristics of the R&D progress data, and the statistical information value is used to represent the attribute information of the target data; the first processing unit 602 processes the data in the first target data set according to the preset algorithm 1 and the autoencoder to obtain a second target data set, wherein the fluctuation characteristics of the data in the second target data set are stronger than the fluctuation characteristics of the data in the first target data set, and the autoencoder is used to strengthen the fluctuation characteristics of the data in the first target data set when encoding the data in the first target data set; the first prediction unit 603 predicts anomalies in the R&D progress according to the preset algorithm 2 and the second target data set, which solves the problem that it is difficult to make anomaly warnings in advance when predicting anomalies in the R&D progress by setting thresholds in the related art, resulting in low R&D efficiency. By processing the data in the acquired first target data set to obtain the second target data set, and predicting anomalies in the R&D progress according to the preset algorithm 2 and the second target data set, anomalies in the R&D progress can be warned in advance, thereby achieving the effect of improving R&D efficiency.

[0078] Optionally, in the abnormality prediction device for R&D progress provided in an embodiment of the present application, the device also includes: a second acquisition unit, used to acquire a third target data set before acquiring the first target data set, wherein the third target data set at least includes the target data; a first calculation unit, used to calculate the statistical information value of the target data based on the third target data set and a sliding window algorithm; and a first combination unit, used to combine the statistical information values ​​of the target data to obtain the first target data set.

[0079] Optionally, in the abnormality prediction device for R&D progress provided in an embodiment of the present application, the device also includes: a third acquisition unit, used to acquire the original data in the original data set before acquiring the third target data set, wherein the original data is data representing the R&D progress; a second calculation unit, used to calculate the L2 norm and the norm change rate of the original data based on the original data and the sliding window algorithm; and a second combination unit, used to combine the L2 norm and the norm change rate to obtain the third target data set.

[0080] Optionally, in the abnormal prediction device for R&D progress provided in the embodiment of the present application, the device also includes: a third calculation unit, used to calculate the amount of work completed in the R&D progress according to the target device before obtaining the original data in the original data set; a fourth calculation unit, used to calculate the R&D progress based on the completed workload and a preset formula; a first conversion unit, used to convert the R&D progress into a vector representation to obtain the original data.

[0081] Optionally, in the abnormal prediction device for R&D progress provided in the embodiment of the present application, the first prediction unit includes: a first determination module, used to determine the abnormal value in the second target data set according to preset algorithm two; a first prediction module, used to make an abnormal prediction of the R&D progress based on the abnormal value.

[0082] Optionally, in the abnormality prediction device for R&D progress provided in the embodiment of the present application, the first prediction module includes: a first processing sub-module, used to determine that the possibility of an abnormality in the R&D progress is greater than a first preset probability if the difference between the abnormal value and the preset value is greater than the first preset value; and a second processing sub-module, used to determine that the possibility of an abnormality in the R&D progress is less than a second preset probability if the difference between the abnormal value and the preset value is less than a second preset value.

[0083] The abnormality prediction device for R&D progress includes a processor and a memory. The above-mentioned first acquisition unit 601, first processing unit 602 and first prediction unit 603 are all stored in the memory as program units, and the processor executes the above-mentioned program units stored in the memory to realize corresponding functions.

[0084] The processor contains a kernel, which calls the corresponding program unit from the memory. One or more kernels can be set, and the R&D efficiency can be improved by adjusting the kernel parameters.

[0085] The memory may include non-permanent memory in a computer-readable medium, random access memory (RAM) and / or non-volatile memory, such as read-only memory (ROM) or flash RAM, and the memory includes at least one memory chip.

[0086] An embodiment of the present invention provides a computer-readable storage medium having a program stored thereon, which implements the abnormality prediction method of the R&D progress when the program is executed by a processor.

[0087] An embodiment of the present invention provides a processor, which is used to run a program, wherein the abnormality prediction method of the research and development progress is executed when the program is running.

[0088] like Figure 7As shown, an embodiment of the present invention provides a device, which includes a processor, a memory, and a program stored in the memory and executable on the processor. When the processor executes the program, the following steps are implemented: obtaining a first target data set, wherein the first target data set at least includes statistical information values ​​of target data, wherein the target data is used to represent the fluctuation characteristics of R&D progress data, and the statistical information values ​​are used to represent the attribute information of the target data; processing the data in the first target data set according to a preset algorithm one and an automatic encoder to obtain a second target data set, wherein the fluctuation characteristics of the data in the second target data set are stronger than the fluctuation characteristics of the data in the first target data set, and the automatic encoder is used to strengthen the fluctuation characteristics of the data in the first target data set when encoding the data in the first target data set; performing abnormal prediction on the R&D progress according to a preset algorithm two and the second target data set.

[0089] When the processor executes the program, the following steps are also implemented: before obtaining the first target data set, the method also includes: obtaining a third target data set, wherein the third target data set at least includes the target data; calculating the statistical information value of the target data based on the third target data set and the sliding window algorithm; and combining the statistical information values ​​of the target data to obtain the first target data set.

[0090] When the processor executes the program, the following steps are also implemented: before obtaining the third target data set, the method also includes: obtaining original data in the original data set, wherein the original data is data representing the research and development progress; based on the original data and the sliding window algorithm, calculating the L2 norm and the norm change rate of the original data; combining the L2 norm and the norm change rate to obtain the third target data set.

[0091] When the processor executes the program, the following steps are also implemented: before obtaining the original data in the original data set, the method also includes: calculating the amount of work completed in the research and development progress according to the target device; calculating the research and development progress according to the completed work and a preset formula; converting the research and development progress into a vector representation to obtain the original data.

[0092] When the processor executes the program, the following steps are also implemented: according to the preset algorithm 2 and the second target data set, an abnormal prediction of the research and development progress is made, including: according to the preset algorithm 2, determining the abnormal value in the second target data set; and making an abnormal prediction of the research and development progress based on the abnormal value.

[0093] When the processor executes the program, the following steps are also implemented: performing abnormal prediction on the R&D progress according to the abnormal value includes: if the difference between the abnormal value and the preset value is greater than the first preset value, then determining that the possibility of abnormality in the R&D progress is greater than the first preset probability; if the difference between the abnormal value and the preset value is less than the second preset value, then determining that the possibility of abnormality in the R&D progress is less than the second preset probability. The device in this article can be a server, PC, PAD, mobile phone, etc.

[0094] The present application also provides a computer program product, which, when executed on a data processing device, is suitable for executing a program that is initialized with the following method steps: obtaining a first target data set, wherein the first target data set at least includes statistical information values ​​of target data, wherein the target data is used to represent the fluctuation characteristics of research and development progress data, and the statistical information values ​​are used to represent attribute information of the target data; processing the data in the first target data set according to a preset algorithm one and an automatic encoder to obtain a second target data set, wherein the fluctuation characteristics of the data in the second target data set are stronger than the fluctuation characteristics of the data in the first target data set, and the automatic encoder is used to strengthen the fluctuation characteristics of the data in the first target data set when encoding the data in the first target data set; performing abnormal prediction of the research and development progress according to a preset algorithm two and the second target data set.

[0095] When executed on a data processing device, it is also suitable for executing a program initialized with the following method steps: before obtaining the first target data set, the method also includes: obtaining a third target data set, wherein the third target data set at least includes the target data; calculating the statistical information value of the target data based on the third target data set and a sliding window algorithm; and combining the statistical information values ​​of the target data to obtain the first target data set.

[0096] When executed on a data processing device, it is also suitable for executing a program initialized with the following method steps: before obtaining the third target data set, the method also includes: obtaining original data in the original data set, wherein the original data is data representing the research and development progress; based on the original data and the sliding window algorithm, calculating the L2 norm and the norm change rate of the original data; combining the L2 norm and the norm change rate to obtain the third target data set.

[0097] When executed on a data processing device, it is also suitable for executing a program initialized with the following method steps: before obtaining the original data in the original data set, the method also includes: calculating the amount of work completed in the research and development progress according to the target device; calculating the research and development progress according to the completed amount of work and a preset formula; converting the research and development progress into a vector representation to obtain the original data.

[0098] When executed on a data processing device, it is also suitable for executing a program initialized with the following method steps: performing an abnormal prediction on the research and development progress according to preset algorithm two and the second target data set, including: determining the abnormal value in the second target data set according to the preset algorithm two; performing an abnormal prediction on the research and development progress based on the abnormal value.

[0099] When executed on a data processing device, it is also suitable for executing a program initialized with the following method steps: making an abnormal prediction of the research and development progress based on the abnormal value includes: if the difference between the abnormal value and the preset value is greater than a first preset value, then determining that the possibility of an abnormality in the research and development progress is greater than a first preset probability; if the difference between the abnormal value and the preset value is less than a second preset value, then determining that the possibility of an abnormality in the research and development progress is less than a second preset probability.

[0100] Those skilled in the art will appreciate that the embodiments of the present application may be provided as methods, systems, or computer program products. Therefore, the present application may adopt the form of a complete hardware embodiment, a complete software embodiment, or an embodiment in combination with software and hardware. Moreover, the present application may adopt the form of a computer program product implemented in one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) that include computer-usable program code.

[0101] The present application is described with reference to the flowcharts and / or block diagrams of the methods, devices (systems), and computer program products according to the embodiments of the present application. It should be understood that each process and / or box in the flowchart and / or block diagram, as well as the combination of the processes and / or boxes in the flowchart and / or block diagram, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing device to generate a machine, so that the instructions executed by the processor of the computer or other programmable data processing device generate instructions for implementing the processes in the flowchart and / or block diagram. Figure 1 A process or multiple processes and / or boxes Figure 1 A device that provides the functions specified in a block or multiple blocks.

[0102] These computer program instructions may also be stored in a computer-readable memory capable of directing a computer or other programmable data processing device to operate in a specific manner, so that the instructions stored in the computer-readable memory produce an article of manufacture comprising an instruction device, which implements the process Figure 1 A process or multiple processes and / or boxes Figure 1 A function specified in one or more boxes.

[0103] These computer program instructions can also be loaded onto a computer or other programmable data processing device so that a series of operating steps are executed on the computer or other programmable device to produce a computer-implemented process, thereby providing instructions for implementing the process. Figure 1 A process or multiple processes and / or boxes Figure 1 The steps for the functions specified in one or more boxes.

[0104] In a typical configuration, a computing device includes one or more processors (CPU), input / output interfaces, network interfaces, and memory.

[0105] The memory may include non-permanent memory in a computer-readable medium, random access memory (RAM) and / or non-volatile memory in the form of read-only memory (ROM) or flash RAM. The memory is an example of a computer-readable medium.

[0106] Computer readable media include permanent and non-permanent, removable and non-removable media that can be implemented by any method or technology to store information. Information can be computer readable instructions, data structures, program modules or other data. Examples of computer storage media include, but are not limited to, phase change memory (PRAM), static random access memory (SRAM), dynamic random access memory (DRAM), other types of random access memory (RAM), read-only memory (ROM), electrically erasable programmable read-only memory (EEPROM), flash memory or other memory technology, compact disk read-only memory (CD-ROM), digital versatile disk (DVD) or other optical storage, magnetic cassettes, magnetic tape magnetic disk storage or other magnetic storage devices or any other non-transmission media that can be used to store information that can be accessed by a computing device. As defined herein, computer readable media does not include temporary computer readable media (transitory media), such as modulated data signals and carrier waves.

[0107] It should also be noted that the terms "include", "comprises" or any other variations thereof are intended to cover non-exclusive inclusion, so that a process, method, commodity or device including a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such process, method, commodity or device. In the absence of more restrictions, the elements defined by the sentence "comprises a ..." do not exclude the existence of other identical elements in the process, method, commodity or device including the elements.

[0108] Those skilled in the art will appreciate that the embodiments of the present application may be provided as methods, systems or computer program products. Therefore, the present application may adopt the form of a complete hardware embodiment, a complete software embodiment or an embodiment in combination with software and hardware. Moreover, the present application may adopt the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) that contain computer-usable program code.

[0109] The above are only embodiments of the present application and are not intended to limit the present application. For those skilled in the art, the present application may have various changes and variations. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present application should be included within the scope of the claims of the present application.

Claims

1. A method for predicting abnormality of R&D progress, characterized in that: include: Acquire a first target data set, wherein the first target data set at least includes statistical information values ​​of target data, wherein the target data is used to represent the fluctuation characteristics of the R&D progress data, and the statistical information values ​​are used to represent the attribute information of the target data, and the statistical information values ​​include: mean, standard deviation, minimum value, maximum value, quartile and range; According to a preset algorithm 1 and an autoencoder, the data in the first target data set is processed to obtain a second target data set, wherein the fluctuation characteristics of the data in the second target data set are stronger than the fluctuation characteristics of the data in the first target data set, and the autoencoder is used to strengthen the fluctuation characteristics of the data in the first target data set when encoding the data in the first target data set, and the preset algorithm 1 is one of the following: a long short-term memory network algorithm, a gated recurrent unit algorithm, and a recurrent neural network algorithm; According to a second preset algorithm and the second target data set, an abnormal prediction is performed on the research and development progress, wherein the second preset algorithm is an isolation forest algorithm or a random forest algorithm; Before acquiring the first target data set, the method further includes: Acquire a third target data set, wherein the third target data set at least includes the target data; Calculating a statistical information value of the target data according to the third target data set and a sliding window algorithm; Combining the statistical information values ​​of the target data to obtain the first target data set; Before acquiring the third target data set, the method further includes: Obtaining original data in an original data set, wherein the original data is data representing the research and development progress; Calculate the L2 norm and the norm change rate of the original data according to the original data and the sliding window algorithm; The L2 norm and the norm change rate are combined to obtain the third target data set.

2. The method according to claim 1, characterized in that Before obtaining the original data in the original data set, the method further includes: Calculate the amount of work completed in the R&D schedule based on the target device; Calculate the research and development progress based on the completed workload and a preset formula; The research and development progress is converted into a vector representation to obtain the original data.

3. The method according to claim 1, characterized in that According to the preset algorithm 2 and the second target data set, abnormal prediction of the research and development progress includes: According to the preset algorithm 2, determining abnormal values ​​in the second target data set; An abnormal prediction is made on the research and development progress based on the abnormal value.

4. The method according to claim 3, characterized in that The abnormal prediction of the research and development progress according to the abnormal value includes: If the difference between the abnormal value and the preset value is greater than a first preset value, it is determined that the possibility that the research and development progress is abnormal is greater than a first preset probability; If the difference between the abnormal value and the preset value is smaller than a second preset value, it is determined that the possibility that the research and development progress is abnormal is smaller than a second preset probability.

5. A device for predicting abnormality of R&D progress, characterized in that: include: A first acquisition unit is used to acquire a first target data set, wherein the first target data set at least includes statistical information values ​​of target data, wherein the target data is used to represent the fluctuation characteristics of the R&D progress data, and the statistical information values ​​are used to represent the attribute information of the target data, and the statistical information values ​​include: mean, standard deviation, minimum value, maximum value, quartile and range; A first processing unit is used to process the data in the first target data set according to a preset algorithm 1 and an autoencoder to obtain a second target data set, wherein the fluctuation characteristics of the data in the second target data set are stronger than the fluctuation characteristics of the data in the first target data set, and the autoencoder is used to strengthen the fluctuation characteristics of the data in the first target data set when encoding the data in the first target data set, and the preset algorithm 1 is one of the following: a long short-term memory network algorithm, a gated recurrent unit algorithm, and a recurrent neural network algorithm; A first prediction unit, configured to perform abnormal prediction on the research and development progress according to a second preset algorithm and the second target data set, wherein the second preset algorithm is an isolation forest algorithm or a random forest algorithm; A second acquisition unit, configured to acquire a third target data set before acquiring the first target data set, wherein the third target data set at least includes the target data; A first calculation unit, configured to calculate a statistical information value of the target data according to the third target data set and a sliding window algorithm; A first combining unit, configured to combine the statistical information values ​​of the target data to obtain the first target data set; The device further comprises: a third acquisition unit, configured to acquire original data in the original data set before acquiring the third target data set, wherein the original data is data representing the research and development progress; The second calculation unit is used to calculate the L2 norm and the norm change rate of the original data based on the original data and the sliding window algorithm; the second combination unit is used to combine the L2 norm and the norm change rate to obtain the third target data set.

6. A computer-readable storage medium, characterized in that: The storage medium includes a stored program, wherein the program executes the abnormality prediction method for R&D progress according to any one of claims 1 to 4.

7. An electronic device, characterized in that: It includes one or more processors and a memory, wherein the memory is used to store one or more programs, wherein when the one or more programs are executed by the one or more processors, the one or more processors implement the abnormality prediction method of the R&D progress described in any one of claims 1 to 4.

Images