KR-20260065011-A - METHOD AND DEVICE FOR PREDICTING THE DURATION OF HOSPITAL STAFF WORK

KR20260065011AKR 20260065011 AKR20260065011 AKR 20260065011AKR-20260065011-A

Abstract

An embodiment of the present invention relates to a method and apparatus for predicting the working period of hospital personnel, and more specifically, to a method and apparatus for predicting the working period of hospital personnel capable of providing a service for deriving the predicted working period of personnel who are currently working or are scheduled to work at a hospital.

Inventors

고승현
정연웅
조재형
장진환
장흥수
심진보

Assignees

가톨릭대학교 산학협력단

Dates

Publication Date: 20260508
Application Date: 20241031

Claims (14)

As a method for calculating the expected employment period of hospital staff, Step of collecting personal and work data of existing personnel working at the hospital, A step of generating normalized data by preprocessing the personal data and work data of the aforementioned existing personnel, A step of setting weights for each item included in the personal data and each item included in the work data based on the normalized data above. A step of generating a hospital staff employment period prediction model trained to derive the expected employment period of hospital staff using the above weights, and A step of inputting the personal data and work history data of hospital personnel into the hospital personnel work period prediction model to derive the expected work period of the hospital personnel. A method for calculating the expected working period of hospital personnel, including
In paragraph 1, The step of collecting personal data and work data of the aforementioned existing personnel is, A step of setting the age, gender, educational background, certifications, work experience, major, and number of job changes of the aforementioned existing personnel as items included in personal data, and Step of setting the work department, working hours, work type, hospital size, hospital location, average commute distance, and annual vacation days of the aforementioned existing personnel as items included in the work data A method for calculating the expected working period of hospital personnel, including
In paragraph 2, The step of generating the above normalized data is, A step of generating a distribution graph for each item included in the personal data of the existing personnel and the work data of the existing personnel, A step of generating the normalized data by removing noise data included in the distribution graph based on the normal distribution graph. A method for calculating the expected working period of hospital personnel, including
In paragraph 3, The step of setting weights for each of the above items is, A step of generating item-by-item graphs of the personal data of the existing personnel and the work data of the existing personnel, A step of deriving the agreement rate by comparing the above item-specific graphs with a normal distribution graph, A step of deriving a first weight for each item based on the above matching rate, A step of deriving item-specific gradients for a loss function and deriving a second weight based on the said gradients, A step of setting the average value of the first weight and the second weight as the final weight, and Step of generating a hospital workforce employment period prediction model retrained based on the above final weights A method for calculating the expected working period of hospital personnel, including
In paragraph 4, The step of deriving the first weight above is, Step of deriving the first weight using the first mathematical formula Includes, [First Mathematical Formula] In the above mathematical formula 1 is an item The distribution function of the actual data for, is the normal distribution function, is the standard deviation between two distributions, The item A method for calculating the expected working period of hospital personnel, which signifies the first weighting for
In paragraph 5, The step of deriving the second weight mentioned above is, Step of deriving the second weight using the second mathematical formula Includes, [Second Mathematical Formula] In the second mathematical formula above is the loss function of the above hospital staff employment period prediction model, Key items affecting employment period , is the relevant item A method for calculating the expected working period of hospital personnel, meaning a second weight for
In paragraph 6, The step of generating the above-mentioned learned hospital workforce employment period prediction model is, A step of generating a hospital staff employment duration prediction model to include one or more of Large Language Models (LLM), Random Forest, Gradient Boosting, or Multilayer Perceptron. A method for calculating the expected working period of hospital personnel, including
As a device that provides a service for calculating the expected working period of hospital personnel, Memory for storing the program to calculate the estimated working period of hospital staff, and It includes a processor that executes a program for calculating the estimated working period of hospital personnel stored in the memory above, and The above processor is, A device for providing a service for calculating the expected working period of hospital personnel, which collects personal data and work data of existing personnel working at a hospital, preprocesses the personal data and work data of the existing personnel to generate normalized data, sets weights for each item included in the personal data and each item included in the work data based on the normalized data, generates a hospital personnel working period prediction model trained to derive the expected working period of the hospital personnel using the weights, and inputs the personal data and work history data of the hospital personnel into the hospital personnel working period prediction model to derive the expected working period of the hospital personnel.
In paragraph 8, The above processor is, A device for providing a service to calculate the expected working period of hospital personnel, which sets the age, gender, educational background, certifications, career, major, and number of job changes of the aforementioned existing personnel as items included in personal data, and sets the work department, working hours, work type, hospital size, hospital location, average commuting distance, and annual vacation days of the aforementioned existing personnel as items included in work data.
In Paragraph 9, The above processor is, A device for providing a service for calculating the expected working period of hospital personnel, which generates a distribution graph for each item included in the personal data and work data of the existing personnel, and generates the normalized data by removing noise data included in the distribution graph based on the normal distribution graph.
In Paragraph 10, The above processor is, A device for providing a service for calculating the expected working period of hospital personnel, comprising: generating item-specific graphs of the personal data of the existing personnel and the work data of the existing personnel; deriving a match rate by comparing the item-specific graphs with a normal distribution graph; deriving a first weight for each item based on the match rate; deriving an item-specific slope for a loss function; deriving a second weight based on the slope; setting the average value of the first weight and the second weight as the final weight; and generating a retrained hospital personnel working period prediction model based on the final weight.
In Paragraph 11, The above processor is, The first weight is derived using mathematical formula 1, and [Mathematical Formula 1] In the above mathematical formula 1 is an item The distribution function of the actual data for, is the normal distribution function, is the standard deviation between two distributions, The item A service providing device for calculating the expected working period of hospital personnel, representing the first weighting for
In Paragraph 12, The above processor is, The above second weight is derived using mathematical formula 2, and [Mathematical Formula 2] In the above mathematical formula 2 is the loss function of the above hospital staff employment period prediction model, Key items affecting employment period , is the relevant item A service providing device for calculating the expected working period of hospital personnel, signifying a second weighting factor.
In Paragraph 13, The above processor is, A device providing a service for calculating the expected working period of hospital personnel, which generates a hospital personnel working period prediction model including one or more of Large Language Models (LLM), Random Forest, Gradient Boosting, or Multilayer Perceptron.

Description

Method and Device for Predicting the Duration of Hospital Staff Work The present invention relates to a method and apparatus for predicting the employment period of hospital personnel, and more specifically, to a method and apparatus for predicting the employment period of hospital personnel based on information of personnel working at the hospital or applicants who have applied for hospital employment. In medical institutions such as hospitals, various personnel, including doctors and nurses, play crucial roles in patient care and hospital operations. However, staff stability and long-term employment are essential for maintaining operational efficiency and providing high-quality medical services to patients. In particular, staff turnover directly impacts operational efficiency and patient satisfaction, requiring hospitals to invest significant costs and resources in recruiting and training new personnel. Currently, hospitals are making various attempts to predict the tenure of their workforce, and statistical analysis methods are generally used for this purpose. These methods primarily analyze key characteristics that can influence tenure based on existing personnel data and predict the tenure of new hires based on this analysis. However, existing statistical methods have limitations in comprehensively reflecting various factors such as hospital size, work department, and individual characteristics. Furthermore, due to a lack of adaptability to new work environments or unpredictable factors, the accuracy of employment duration predictions may consequently decrease. In particular, at medical institutions, the tenure of staff can vary depending on the characteristics of each department or the size of the hospital. For instance, personnel working in high-intensity departments such as emergency rooms and intensive care units may exhibit relatively higher turnover rates compared to other departments, and differences in working environments between large and small-to-medium-sized hospitals can also affect the length of employment. Recently, machine learning technology is being effectively utilized for data-driven forecasting across various industries, and there is an increasing number of attempts by hospitals to adopt this technology to predict the tenure of their workforce. Compared to traditional statistical analysis methods, machine learning models have the advantage of being able to process multidimensional and complex data and automatically learn key characteristics that influence tenure within a structure where various factors interact. Furthermore, machine learning models have the potential to provide more accurate predictions when forecasting the tenure of new hires based on historical data. However, existing machine learning-based methods are limited to merely extracting data patterns during the training process and often fail to subdivide the impact of each characteristic on working duration or effectively reflect the relationships between characteristics. Furthermore, there is a problem in that the model's predictive performance may degrade if the training weights are not sufficiently linked to the data distribution or statistical characteristics of each characteristic. FIG. 1 is an exemplary diagram showing the communication connection of a hospital staff work period prediction device according to an embodiment of the present invention. Figure 2 is a configuration diagram of a hospital staffing period prediction server according to an embodiment of the present invention. FIG. 3 is a configuration diagram of a terminal according to an embodiment of the present invention. FIG. 4 is a conceptual diagram illustrating the functions of a processor according to an embodiment of the present invention. FIGS. 5 and 6 are flowcharts of a method for predicting the working period of hospital personnel according to an embodiment of the present invention. Embodiments of the present invention are described below with reference to the attached drawings so that those skilled in the art can easily implement the invention. However, the present invention may be embodied in various different forms and is not limited to the embodiments described herein. Furthermore, in order to clearly explain the present invention in the drawings, parts unrelated to the explanation have been omitted, and similar parts throughout the specification are denoted by similar reference numerals. Throughout the specification, when a part is described as being "connected" to another part, this includes not only cases where they are "directly connected," but also cases where they are "electrically connected" with other components interposed between them. Furthermore, when a part is described as "including" a certain component, this means that, unless specifically stated otherwise, it does not exclude other components but may include additional components. In addition, the attached drawings are intended only to facilitate understanding of the embodi