Holstein cattle production life estimation method and system based on dynamic regression model
Through the Holstein cattle production lifespan estimation method based on the dynamic regression model, the lag and inaccuracy problems of Holstein cattle lifespan prediction are solved, dynamic prediction at different age stages is achieved, and the accuracy of dairy cow production lifespan and the economic benefits of the ranch are improved.
Patent Information
- Application Number
- CN202510923779.2
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2025-07-04
- Publication Date
- 2025-10-10
AI Technical Summary
The existing Holstein cattle lifespan estimation method has lags, cannot dynamically evaluate lactation potential, lacks consideration of multiple parities, and cannot flexibly update predicted values at different age stages, resulting in inaccurate predictions of dairy cow lifespan, affecting breeding and ranch economic benefits.
A method based on a dynamic regression model was adopted. By collecting individual data of Holstein cows, a multi-parity accumulation mechanism was constructed, and the regression coefficient was dynamically updated. The cumulative number of lactation months for individuals surviving to different age stages was estimated. Combined with physiological and environmental variables, a linear regression model was constructed to achieve early prediction of the productive lifespan of dairy cows.
The accuracy of dairy cow life prediction and the economic benefits of pastures are improved. By screening high-yielding and long-lived dairy cows at an early stage and adjusting feeding and breeding plans, the production life of dairy cows can be extended and the efficiency of genetic selection can be improved.
Smart Images

Figure CN120763899A_ABST
Abstract
Description
TECHNICAL FIELD
[0001] The present application relates to the field of dairy cattle breeding technology, and particularly relates to a Holstein cow production life estimation method and system based on a dynamic regression model. BACKGROUND
[0002] The statements in this section merely provide background information related to the present application and do not necessarily constitute the prior art.
[0003] The life span of a dairy cow such as a Holstein cow is a decisive factor affecting the profitability of a dairy farm, because the cost of raising replacement heifers is high, and only when the cow gives birth to a calf and continues to produce can the farm bring higher benefits, that is, the longer the production life of the cow, the higher the efficiency of the farm. By extending the life of the cow through passive culling, the health cost will be reduced, the lifetime profitability of the cow will be improved, the animal welfare will be improved, and the overall utilization efficiency of the available resources of the cow farm will be optimized. However, the life span data of a cow is special, because only after the cow is culled (or dies) can the real production life of the cow itself be obtained, so early prediction of the production life of the cow is crucial to the economic efficiency of the farm and genetic selection. However, the life span record of a cow does not conform to a normal distribution, and assuming that the current life span of a surviving individual is the real life span will underestimate the life span data, because the life span of a surviving individual must be greater than or equal to its current life span; in addition, in breeding work, the current method is to select a bull with high breeding value to improve the cow, and breeders also want to select a bull with high longevity characteristics as early as possible, and it is obviously too late to evaluate the breeding value of the bull after all its offspring are culled, so it is necessary to estimate the production life breeding value (EBV) of the offspring of the bull as early as possible.
[0004] The existing Holstein cow life estimation method still has some problems: on the one hand, it is mostly an estimation of the life span of an individual, which relies on historical culling data (i.e., individual death data), and has a certain lag; on the other hand, the traditional life estimation method mostly estimates the actual life span or actual survival months of a cow, and the definition of this life trait (such as actual survival months) does not take into account the cumulative effect of lactation efficiency, nor does it take into account the multiple lactation restrictions, and it cannot dynamically evaluate the lactation potential of an individual during its survival period; in addition, the existing method also lacks a dynamic prediction model, which cannot flexibly update the prediction value at different age stages and identify key influencing factors. SUMMARY
[0005] To address the deficiencies of the above-mentioned prior art, the present invention provides a method and system for estimating the productive lifespan of Holstein cattle based on a dynamic regression model, proposes the productive lifespan trait MIM66, which is the cumulative number of lactation months when an individual survives to 66 months of age, designs a multiple-parity accumulation mechanism, and constructs a dynamic regression model to achieve dynamic and accurate estimation of the productive lifespan trait MIM66 of Holstein cattle at different age stages based on multi-stage lactation trait data as well as physiological and environmental variable data, thereby providing data support for genetic selection and breeding of dairy cows and improving the economic benefits of pastures.
[0006] In a first aspect, the present invention provides a method for estimating the productive life span of Holstein cattle based on a dynamic regression model.
[0007] A method for estimating the productive life span of Holstein cattle based on a dynamic regression model includes: Collect data on several Holstein cattle individuals, select data on multiple Holstein cattle individuals with a lifespan of 66 months, and construct a data set; Lactation-related variable data, physiological and environmental variable data, and target trait MIM66 data of individuals surviving to different age points were extracted from the data set, and the extracted data were encoded to form a training set; A linear regression model was constructed with lactation-related variables, physiological and environmental variables as independent variables and the target trait MIM66 as the dependent variable. The model was trained using the training set and the regression coefficient was dynamically updated. Obtain the individual data of Holstein cattle to be estimated, and use the dynamic regression model to estimate the productive life trait MIM66 of the individual.
[0008] In a second aspect, the present invention provides a system for estimating the productive life span of Holstein cattle based on a dynamic regression model.
[0009] A Holstein cattle production life estimation system based on a dynamic regression model, comprising: A data collection module is used to collect data on several Holstein cattle individuals, screen data on multiple Holstein cattle individuals with a lifespan of 66 months, and construct a data set; The data processing module is used to extract lactation-related variable data, physiological and environmental variable data, and target trait MIM66 data of individuals surviving to different age points from the data set, and encode the extracted data to form a training set; The model building and training module is used to build a linear regression model with lactation-related variables, physiological and environmental variables as independent variables and the target trait MIM66 as the dependent variable, train the model using the training set, and dynamically update the regression coefficient; The production life estimation module is used to obtain the individual data of Holstein cattle to be estimated and estimate the production life trait MIM66 of the individual using a dynamic regression model.
[0010] In a third aspect, the present invention further provides an electronic device comprising: a memory for storing executable instructions; and a processor for implementing the above-mentioned method for estimating the production life span of Holstein cattle based on a dynamic regression model when executing the executable instructions stored in the memory.
[0011] In a fourth aspect, the present invention further provides a computer-readable storage medium storing executable instructions for causing a processor to execute the executable instructions to implement the above-mentioned method for estimating the production life span of Holstein cattle based on a dynamic regression model.
[0012] In a fifth aspect, the present invention also provides a computer program product, which includes executable instructions, and the executable instructions are stored in a computer-readable storage medium; wherein, when the processor of the electronic device reads the executable instructions from the computer-readable storage medium and executes the executable instructions, the above-mentioned method for estimating the production life span of Holstein cattle based on the dynamic regression model is implemented.
[0013] One or more of the above technical solutions have the following beneficial effects: The present invention proposes a method and system for estimating the productive lifespan of Holstein cattle based on a dynamic regression model, proposes the productive lifespan trait MIM66, which is the cumulative number of lactation months when an individual survives to 66 months of age, designs a multiple-parity accumulation mechanism, and constructs a dynamic regression model. The method can dynamically and accurately estimate the productive lifespan trait MIM66 of Holstein cattle at different age stages based on multi-stage lactation trait data as well as physiological and environmental variable data, providing data support for genetic selection of dairy cows and improving the economic benefits of pastures. For example, high-yield and long-lived dairy cows can be selected as a core group based on early predicted MIM66 values, and individuals with low MIM66 predicted values can be intervened in advance to adjust feeding or breeding plans, so as to further improve individual productive lifespan and improve the economic benefits of pastures.
[0014] Advantages of additional aspects of the present invention will be given in part in the following description and in part will be obvious from the following description, or will be learned through practice of the present invention. BRIEF DESCRIPTION OF THE DRAWINGS
[0015] The accompanying drawings, which constitute a part of the present invention, are used to provide a further understanding of the present invention. The exemplary embodiments of the present invention and their descriptions are used to explain the present invention and do not constitute improper limitations on the present invention.
[0016] Figure 1 This is a flow chart of a method for estimating the productive lifespan of Holstein cattle based on a dynamic regression model in an embodiment of the present invention; Figure 2 Schematic diagram of the production life characteristics of MIM66 in an embodiment of the present invention; Figure 3This is a schematic diagram showing the comparison between the predicted value and the actual value of estimating the productive life span of individual dairy cows using the constructed model in an embodiment of the present invention. DETAILED DESCRIPTION
[0017] It should be noted that the following detailed descriptions are exemplary only and are intended to describe specific embodiments and provide further explanation of the present invention, and are not intended to limit the exemplary embodiments according to the present invention. Unless otherwise indicated, all technical and scientific terms used herein have the same meanings as those commonly understood by those of ordinary skill in the art to which the present invention belongs. In addition, it should be understood that when the terms "comprise" and / or "include" are used in this specification, they indicate the presence of features, steps, operations, devices, components and / or combinations thereof.
[0018] Example 1 This embodiment provides a method for estimating the productive life span of Holstein cattle based on a dynamic regression model. Figure 1 As shown, the specific steps include: Step S1: Collect individual data of several Holstein cattle, select individual data of multiple Holstein cattle with a lifespan of up to 66 months, and construct a data set.
[0019] Specifically, data on several Holstein cows on the ranch, including lactation records, reproductive records, milk production data, and environmental data, were collected from the ranch management system and the dairy performance laboratory (DHI). Furthermore, missing values were processed for the collected individual data. For example, multiple imputation was used to fill in variables such as age at first calving (AFC) and first parity somatic cell score (SCS). Invalid records, such as those with missing sire and dam numbers and those with discontinuous parities, were deleted.
[0020] In this embodiment, a total of 47,312 Holstein cattle individual data were collected, and 4,841 individuals with a lifespan of 66 months were finally selected (all of these individuals had real MIM66) for regression model training. Compared with the traditional individual lifespan estimation method, the training data used in this embodiment is the individual data with a lifespan of 66 months, rather than the individual elimination data, which can avoid the hysteresis that exists when only the elimination data is used to train the model. Moreover, the regression model trained using this data in this embodiment is used to estimate the cumulative number of lactation months when the individual survives to 66 months of age, rather than directly estimating the individual lifespan. The MIM66 value predicted by the individual in the early stage is used to screen high-yield and long-lived dairy cows as the core group, and the individuals with low MIM66 prediction values are intervened in advance to adjust the feeding or breeding plan to further improve the individual production lifespan and improve the economic benefits of the ranch.
[0021] Step S2: extracting lactation-related variable data, physiological and environmental variable data, and target trait MIM66 data of individuals surviving to different age points from the data set, and encoding the extracted data to form a training set.
[0022] Specifically, this embodiment first defines the target trait MIM66, such as Figure 2 As shown in the figure, if a dairy cow survives to 66 months of age, MIM66 is the sum of the lactation months for each parity that survived to 66 months. The maximum lactation month for a single parity is 10 months; any months beyond that are dry months and not lactation months. Furthermore, if a dairy cow is culled early, MIM66 is the actual cumulative lactation months before culling. It should be noted that this example uses data from individuals that have survived to 66 months of age for model training, and does not use data from culled cows.
[0023] Secondly, dynamic prediction variables are defined, including lactation-related variables, physiological and environmental variables. Among them, lactation-related variables include cumulative lactation months, current parity status, current parity lactation months, and current parity dry months, specifically: Cumulative months: This variable is a continuous variable, which refers to the cumulative number of lactation months for each individual when they survive to a specific lifespan age point. The number of months is accumulated by parity, and each parity does not exceed 10 months. Among them, the specific lifespan age point refers to the set lifespan time points such as 30 months, 36 months, 42 months, 44 months, 46 months, 48 months, 54 months, and 60 months.
[0024] Regarding the current parity status (lactation status): This variable is a categorical variable, which refers to the parity status of each individual when it survives to a specific life age point. When the individual is in the "dry month" state, the status is 0, and when the individual is in the "lactation month" state, the status is 1. Among them, the number of lactation months for each parity does not exceed 10 months. When the individual survives to a specific life age point, the number of lactation months for the current parity has exceeded the set number of lactation months (i.e., 10 months), then the current status of the individual is the "dry month" state. Correspondingly, when the individual survives to a specific life age point, the number of lactation months for the current parity has not exceeded the set number of lactation months (i.e., 10 months), then the current status of the individual is the "lactation month" state. For example, when an individual reaches 36 months of age, the fetus has been lactating for 320 days, and the current lactation days exceed 10 months (assuming an average of 30.5 days per month), and the individual is in the "dry month" state, that is, the state is 0; if the fetus has been lactating for 150 days, the current lactation days do not exceed 10 months, and the individual is in the "lactation month" state, that is, the state is 1.
[0025] Current months of lactation: This variable is a continuous variable, which refers to the number of lactation months of each individual's parity when they survive to a specific life span age point, with a maximum of 10 months.
[0026] Regarding the current parity dry months variable (current months dry): This variable is a continuous variable and refers to the number of dry months for each parity when the individual survives to a specific lifespan age point, that is, the number of months that the current parity exceeds the lactation months (10 months). Typically, the interval between parities for dairy cows is approximately 390 days, or 13 months. This means that the interval from the start of lactation in one parity to the start of the next often exceeds 10 months. Therefore, the period of lactation exceeding 10 months in the current parity is considered the "dry months." In practice, this current parity dry months variable is related to the current parity status. If the current parity status is 1, meaning it has not exceeded 10 months, the current parity dry months variable is 0. By setting this current parity dry months variable, it reflects the individual's reproductive status. If an individual suffers from reproductive problems, the interval between parities may be longer, which ultimately affects the cumulative lactation months at 66 months of age, thereby more accurately reflecting the target trait MIM66.
[0027] Among them, physiological and environmental variables include farm-year-season HYS, age at first calving AFC, first calving season SFC, first calving somatic cell score , milk production in the first 305 days F305 Farm-year-season is a combination of fixed effects used to control systematic environmental and management factors that are not individual dairy cows themselves and vary with time, space, and season. It is a combination of the individual's "birth farm, birth year, and birth season"; the age at first calving is the interval between the individual's birth date and the first calving date; the first calving season is the season of the first calving; the first-calving somatic cell score is derived from the DHI laboratory measurement data; and the first-calving milk production at 305 days is derived from the DHI laboratory measurement data.
[0028] Furthermore, lactation-related variable data, physiological and environmental variable data, and target trait MIM66 data of individuals surviving to different age points (such as 30 months, 36 months, 42 months, 44 months, 46 months, 48 months, 54 months, and 60 months) were extracted from the data set, and the extracted data were encoded to form a training set. The data of these individuals with real production lifespans were used to establish subsequent prediction models.
[0029] Preferably, considering that text data may exist in the above-extracted variable data, the extracted data is encoded into numerical form to facilitate the subsequent construction of the model. Taking the farm-year-season as an example, for an individual, the farm number of its birth is 37JN01, the year of birth is 2020, and the season of birth is summer, then the "farm-year-season combination" is "37JN012020 summer", and then the corresponding Python function, such as the pd.factorize() function, is used to convert the variable data into numerical form. Through this data recoding, the text data in the dataset can be converted into numerical data, the category data can be converted into numerical indexes, or the date format can be adjusted to a unified format, etc., so as to meet subsequent analysis needs; for the age at first calving, it is the individual's "calving date of the first parity" minus the "birth date" to obtain the interval number of days, and the number of days is then converted into months, which is the age at first calving; for the first calving season, it is the season in which the individual calved its first parity. In this embodiment, according to the different stress responses of dairy cows to the ambient temperature during calving, the year is divided into two seasons, in which summer and autumn are combined into the hot season, and spring and winter are combined into the cold season, and the two seasons are then represented by 0 and 1 respectively.
[0030] Step S3: Using lactation-related variables, physiological and environmental variables as independent variables and the target trait MIM66 as the dependent variable, a linear regression model is constructed, the model is trained using the training set, and the regression coefficient is dynamically updated.
[0031] Specifically, based on the variable data obtained above, lactation-related variables, physiological and environmental variables were used as independent variables, and the target trait MIM66 was used as the dependent variable to conduct a staged regression modeling, including: First, a generalized linear regression model was established based on the true target trait MIM66 of individual dairy cows surviving to different age points (such as 30, 36, and 48 months of age) as the target variable, which can be expressed as: ; In the above formula, is the intercept term, The cumulative number of lactation months when the lifespan reaches 66 months; Field year and season effect; The effect of first calving age; It is the effect of first calving season; is the cumulative lactation months effect at a specific age point; is the parity status effect; is the effect of lactation months of the current parity; The effect of dry milk months for the current parity; This is the milk production effect at 305 days of first calving; Scoring effects for first-born somatic cells; is the random residual; is the partial regression coefficient.
[0032] Secondly, the regression coefficients are updated dynamically, that is, the model is refitted at different age points using the training set to obtain the coefficient table of each variable at different stages, so as to support real-time prediction and adjustment.
[0033] The physiological and environmental variables introduced in the above model, such as the farm-year-season effect, the first calving age effect, and the first calving season effect, coexist with the lactation-related independent variables and are independent of each other. These effects are introduced as control variables to absorb the systematic variations in the data caused by specific groupings (such as farm-year-season, first calving age group, and first calving season group), thereby separating these variations from the error terms of the model. Under the conditions of stabilizing pasture differences, annual trends, seasonal fluctuations, first calving age, and seasonal differences, the model can then analyze the effects of the core independent variables on lactation performance, which can further improve the accuracy of the model.
[0034] In this embodiment, as shown in Table 1 below, a total of 47,312 Holstein cattle data were collected, of which only 10.2% of the individuals had a lifespan exceeding 66 months. 4,841 individuals with a lifespan of 66 months were selected, and a model was constructed using the statsmodels library of Python. The identity function (dependent variable MIM66 is a continuous positive number) was selected to fit the generalized linear models of these individuals when they survived to 30 months, 36 months, 42 months, 44 months, 46 months, 48 months, 54 months, and 60 months of age. It should be noted that the lifespan of the individuals used for training in this embodiment has reached 66 months of age, so each individual has the value of each variable at each age point such as 36 months and 42 months of age. Therefore, the data set used for training is balanced, and the model obtained by training has better accuracy and universality. The results of the model output are shown in Table 2 below.
[0035] Table 1 Data structure
[0036] Table 2 Model output results
[0037] As an implementation method, the partial regression coefficient of the model is standardized. The standardization formula is: ; in, is the partial regression coefficient after standardization, is the independent variable The partial regression coefficient of is the independent variable The standard deviation of is the dependent variable Y The standard deviation of .
[0038] Through the above standardization processing, we can obtain the partial regression coefficient values under different age estimates as shown in Tables 3 and 4 below.
[0039] Table 3 Partial regression coefficients of the regression model estimating MIM66 at 36 months of age
[0040] Table 4 Partial regression coefficients of the regression model estimating MIM66 at 46 months of age
[0041] As an implementation method, the standardized partial regression coefficients can be used The absolute values of the regression coefficients of different variables are compared to evaluate the impact of different variables on MIM66, and then the variables with greater influence can be selected for prediction and estimation, which can reduce the calculation cost while ensuring the accuracy of the final prediction.
[0042] As another embodiment, the above model can also be expanded to predict the cumulative number of lactation months when surviving to any lifespan, such as MIM84, MIM96, etc., which will not be repeated here.
[0043] Step S4: Obtain the individual data of the Holstein cattle to be estimated, and use the dynamic regression model to estimate the production life trait MIM66 of the individual.
[0044] In this example, the first 30 sample individuals were selected and the dynamic regression model was used to make predictions based on the individual data. The final prediction results are as follows: Figure 3 As shown, by comparing with the real value, the effectiveness and accuracy of the model constructed in this embodiment can be further verified.
[0045] Through the above method proposed in this embodiment, high-yielding and long-lived dairy cows can be selected as a core group based on the early predicted MIM66 values, so as to carry out genetic selection of dairy cows; individuals with low MIM66 predicted values can also be intervened in advance to increase the productive lifespan of dairy cows.
[0046] Example 2 This embodiment provides a system for estimating the productive lifespan of Holstein cattle based on a dynamic regression model, comprising: A data collection module is used to collect data on several Holstein cattle individuals, screen data on multiple Holstein cattle individuals with a lifespan of 66 months, and construct a data set; The data processing module is used to extract lactation-related variable data, physiological and environmental variable data, and target trait MIM66 data of individuals surviving to different age points from the data set, and encode the extracted data to form a training set; The model building and training module is used to build a linear regression model with lactation-related variables, physiological and environmental variables as independent variables and the target trait MIM66 as the dependent variable, train the model using the training set, and dynamically update the regression coefficient; The production life estimation module is used to obtain the individual data of Holstein cattle to be estimated and estimate the production life trait MIM66 of the individual using a dynamic regression model.
[0047] Example 3 This embodiment provides an electronic device, including: a memory for storing executable instructions; and a processor for implementing the above method provided in this embodiment when executing the executable instructions stored in the memory.
[0048] Example 4 This embodiment further provides a computer-readable storage medium storing executable instructions. When the executable instructions are executed by a processor, the processor will be caused to execute the above method provided in this embodiment.
[0049] Example 5 This embodiment provides a computer program product including executable instructions, which are computer instructions stored in a computer-readable storage medium. When a processor of an electronic device reads the executable instructions from the computer-readable storage medium and executes the executable instructions, the electronic device performs the method provided in this embodiment.
[0050] The steps involved in the above embodiments 2 to 5 correspond to those in embodiment 1. For detailed implementation, please refer to the relevant description of embodiment 1. The term "computer-readable storage medium" should be understood as a single medium or multiple media that includes one or more instruction sets; it should also be understood to include any medium that can store, encode, or carry an instruction set for execution by a processor and cause the processor to perform any method of the present invention.
[0051] Those skilled in the art will appreciate that the modules or steps of the present invention described above can be implemented using a general-purpose computer device. Alternatively, they can be implemented using program code executable by a computing device, which can then be stored in a storage device and executed by the computing device. Alternatively, they can be fabricated into separate integrated circuit modules, or multiple modules or steps can be fabricated into a single integrated circuit module for implementation. The present invention is not limited to any specific combination of hardware and software.
[0052] The above description is only a preferred embodiment of the present invention. Although the specific implementation of the present invention is described in conjunction with the accompanying drawings, it does not limit the scope of protection of the present invention. Those skilled in the art should understand that based on the technical solution of the present invention, various modifications or variations that can be made by those skilled in the art without creative work are still within the scope of protection of the present invention.
Claims
1. A method for estimating the productive life span of Holstein cattle based on a dynamic regression model, characterized in that: include: Collect data on several Holstein cattle individuals, select data on multiple Holstein cattle individuals with a lifespan of 66 months, and construct a data set; Lactation-related variable data, physiological and environmental variable data, and target trait MIM66 data of individuals surviving to different age points were extracted from the data set, and the extracted data were encoded to form a training set; A linear regression model was constructed with lactation-related variables, physiological and environmental variables as independent variables and the target trait MIM66 as the dependent variable. The model was trained using the training set and the regression coefficient was dynamically updated. Obtain the individual data of Holstein cattle to be estimated, and use the dynamic regression model to estimate the productive life trait MIM66 of the individual.
2. The method for estimating the productive life span of Holstein cattle based on a dynamic regression model according to claim 1, wherein: The target trait MIM66 refers to the sum of the lactation months of each parity when the individual survives to 66 months of age, or the actual cumulative lactation months before the individual is eliminated; among which, the maximum lactation months for a single parity is 10 months.
3. The method for estimating the productive life span of Holstein cattle based on a dynamic regression model according to claim 1, wherein: The lactation-related variables include the cumulative number of lactation months, the current parity status, the number of lactation months of the current parity, and the number of dry months of the current parity; The cumulative number of lactation months refers to the cumulative number of lactation months for each individual who survives to a specific lifespan age point; the number of months is accumulated by parity, and each parity does not exceed 10 months; The current parity status refers to the parity status of each individual when it survives to a specific lifespan age point. If the number of lactation months of the current parity exceeds the set number of lactation months when the individual survives to the specific lifespan age point, the individual's current status is the dry month status, and the current parity status is 0. If the number of lactation months of the current parity does not exceed the set number of lactation months when the individual survives to the specific lifespan age point, the individual's current status is the lactation month status, and the current parity status is 1. The number of lactation months for the current parity refers to the number of lactation months for each individual when it survives to a specific life span age point; The dry-wean months for the current parity refers to the dry-wean months for each individual when they survive to a specific age point, that is, the number of months that the current parity exceeds the lactation months.
4. The method for estimating the productive life span of Holstein cattle based on a dynamic regression model according to claim 1, wherein: The physiological and environmental variables include farm year and season, age at first calving, first calving season, first parity somatic cell score, and first parity 305-day milk production; farm year and season is the combination of farm, birth year, and birth season; age at first calving is the interval from birth date to first calving date; and first parity calving season is the season when the first calving occurs.
5. The method for estimating the productive life span of Holstein cattle based on a dynamic regression model according to claim 1, wherein: The staged regression modeling is: A linear regression model was established based on the target trait MIM66 of individuals surviving to different age points as the target variable; The training set is used to refit the model at different age points, and the regression system is dynamically updated to obtain the coefficient table of each variable at different stages.
6. The method for estimating the productive life span of Holstein cattle based on a dynamic regression model according to claim 5, wherein: The constructed linear regression model is: ; In the above formula, is the intercept term, The cumulative number of lactation months when the lifespan reaches 66 months; Field year and season effect; The effect of first calving age; It is the effect of first calving season; is the cumulative lactation months effect at a specific age point; It is the parity status effect; is the effect of lactation months of the current parity; The effect of dry milk months for the current parity; This is the milk production effect at 305 days of first calving; Scoring effects for first-born somatic cells; is the random residual; is the partial regression coefficient.
7. A Holstein cattle production life estimation system based on a dynamic regression model, characterized in that: include: A data collection module is used to collect data on several Holstein cattle individuals, screen data on multiple Holstein cattle individuals with a lifespan of 66 months, and construct a data set; The data processing module is used to extract lactation-related variable data, physiological and environmental variable data, and target trait MIM66 data of individuals surviving to different age points from the data set, and encode the extracted data to form a training set; The model building and training module is used to build a linear regression model with lactation-related variables, physiological and environmental variables as independent variables and the target trait MIM66 as the dependent variable, train the model using the training set, and dynamically update the regression coefficient; The production life estimation module is used to obtain the individual data of Holstein cattle to be estimated and estimate the production life trait MIM66 of the individual using a dynamic regression model.
8. An electronic device, characterized in that: include: a memory for storing executable instructions; The processor is configured to implement the method for estimating the production life span of Holstein cattle based on a dynamic regression model as described in any one of claims 1 to 6 when executing the executable instructions stored in the memory.
9. A computer-readable storage medium, characterized in that Executable instructions are stored, which are used to cause a processor to execute the executable instructions to implement the method for estimating the production life span of Holstein cattle based on a dynamic regression model as described in any one of claims 1 to 6.
10. A computer program product, characterized in that The computer program product includes executable instructions stored in a computer-readable storage medium; When the processor of the electronic device reads the executable instructions from the computer-readable storage medium and executes the executable instructions, the method for estimating the production life span of Holstein cattle based on a dynamic regression model as described in any one of claims 1 to 6 is implemented.
Citation Information
Patent Citations
Dairy cow lactation behavior evaluation method and system
CN114742468A
Breeding method for screening high-quality high-yield dairy cows based on measurement day model and animal model BLUP
CN115281145A
Genetic evaluation method for dairy cow hybridization improvement, dairy cow hybridization method and method for optimizing dairy cow production mode
CN118942539A
method for EARLY SELECTION OF COWS FOR PRODUCTIVE LONGEVITY IN A BULL-PRODUCING BREEDING CORE
RU2007141669A
Method of Assessing Productivity of Lactating Animals Using Fitted Parameters to a Mechanistic Lactation Model
US20060271302A1
Cited By
Method and system for early prediction and intelligent breeding of longevity of dairy cow, and computing device
CN121502732A
