Civil going-out detection method and system based on machine learning

Abstract

The invention relates to a caregiver going-out detection method and system based on machine learning. The method comprises the following steps: using a GPS data acquisition system to collect position data of a current work order of a caregiver; cleaning the position data, and filtering false data to obtain preprocessed position data; extracting the preprocessed position data and the historical work order position data of the current work order, and making a decision on the service position of the nursing staff so as to determine whether there is an outgoing behavior; and if it is determined that the caregiver has the going-out behavior, giving an alarm to remind related supervisors. According to the machine learning-based caregiver going-out detection method provided by the invention, the working point location of the caregiver can be automatically and intelligently analyzed by utilizing a clustering algorithm based on a machine learning technology, and the problem that the going-out behavior of the caregiver cannot be judged due to inaccurate GPS positioning data is effectively solved.

Description

Technical Field

[0001] The present invention relates to the technical field of artificial intelligence, and particularly relates to a caregiver going-out detection method and system based on machine learning. Background Art

[0002] With the increase in the demand for medical services and the intensification of the aging population trend, the role of caregivers in the field of healthcare has become increasingly important. They are mainly responsible for providing basic care, rehabilitation care, and daily care services for patients, which has a direct impact on the physical health of patients. Therefore, it is crucial to supervise the standardization of caregivers' behaviors and the quality of services.

[0003] The existing caregiver supervision systems mainly include: 1. Systems based on manual management, which mainly rely on caregivers to punch in manually, fill in work reports, or be regularly inspected by managers. Although this method is simple and easy to implement, it has obvious drawbacks, such as easy data fraud, inability to monitor in real time, information lag, etc. In addition, the workload of manual inspections is large and the efficiency is low. Especially in the case of a large number of caregivers and a wide service area, it is difficult to ensure the effectiveness and timeliness of monitoring. 2. Systems based on punching cards or signing in, such as those mentioned in Patent CN113808293A. This method mainly introduces a punching card and signing-in system on the elderly care service platform. Caregivers need to punch in at the elderly's home at the start and end of the service. Although this method can confirm whether the caregiver has arrived or left the service location, it cannot provide real-time location information during the process, and it is easy for caregivers to punch in at other locations. 3. GPS-based service area supervision systems.

[0004] The existing caregiver supervision systems have certain limitations, mainly manifested as: 1. Lack of real-time performance. Most of the existing systems rely on manual punching or regular inspections and cannot monitor the behaviors of caregivers in real time. For example, the manual signing-in system only records the location when the caregiver arrives or leaves and cannot capture the behaviors during the service. Once the caregiver leaves without permission during the service, the system may not be able to detect it in time. 2. Positioning accuracy problem. Although the GPS-based monitoring system can provide real-time positioning, in some cases (such as indoors or places with poor signals), the GPS positioning accuracy is not high enough, which may lead to false alarms. For example, when a caregiver briefly leaves the room or moves within a relatively close service area, it may be misjudged as leaving the preset service range. This positioning error may affect the accuracy of judgment and increase the false alarm rate of the system. 3. Lack of intelligent analysis. Most of the current systems only rely on simple rule settings, such as setting geographical fences (Geo-fence) or relying on manual judgment, and lack the ability of intelligent anomaly detection. The system cannot learn the rules from historical data and cannot automatically adjust the monitoring strategy, resulting in limited ability to identify abnormal situations. Summary of the Invention

[0005] In view of some or all of the problems in the prior art, the present invention provides a method for detecting the going-out of caregivers based on machine learning. The method includes the following steps:

[0006] Collect the location data of the current work order of the caregiver using a GPS data collection system;

[0007] Clean the location data, filter out false data, so as to obtain preprocessed location data;

[0008] Extract the preprocessed location data and the historical work order location data of the current work order, and make a decision on the service location of the caregiver this time to determine whether there is a going-out behavior; and

[0009] If it is determined that the caregiver has a going-out behavior, issue a warning to remind the relevant supervisors;

[0010] Among them, extracting the preprocessed location data and the historical work order location data of the current work order, and making a decision on the service location of the caregiver this time to determine whether there is a going-out behavior includes:

[0011] Extract the service point data from the historical work order records of the current work order to form a data set D = {d 1 , d 2 ,..., dn}, where n represents the number of work orders in the data set, and di represents the set of geographical locations where the current work order served in the first i work orders;

[0012] Use the DBSCAN algorithm to perform clustering analysis on the data set D to generate a first cluster set C = {C 1 , C 2 ,..., Ck}, where k represents the number of clusters in the cluster set C;

[0013] Obtain the location data dnew of the current work order of the caregiver, merge the data set D to form D′, where D′ = D ∪ {dnew}, and use the DBSCAN algorithm to perform clustering analysis on the data set D′ to generate a second cluster set C′ = {C 1 ′, C 2 ′,..., Ck″}, where k′ represents the number of clusters in the cluster set C′;

[0014] Compare the first cluster set C and the second cluster set C′ to determine whether a new cluster is generated; and

[0015] If a new cluster is generated, it is determined that the caregiver has a going-out behavior. If no new cluster is generated,

[0016] it is determined that the caregiver does not have a going-out behavior.

[0017] Further, the DBSCAN algorithm is used to perform clustering analysis on the dataset D to generate the first cluster set C = {C 1 , C 2 ,..., Ck}, including:

[0018] Initialize the radius parameter ε and the minimum number of points MinPts. The radius parameter ε is used to define the maximum distance of the density neighborhood, and the minimum number of points MinPts is used to define the minimum number of points included in a cluster;

[0019] Determine the core points. If the number of neighborhood points of a point in the dataset D within the radius parameter ε is greater than or equal to the minimum number of points MinPts, then determine this point as a core point; and

[0020] For each core point, all the points within its ε - radius range plus this core point are taken as a cluster. If the neighborhoods of two core points overlap, then the clusters where the two core points are located are merged into one cluster, and the points that do not belong to any cluster are marked as noise points and reserved separately.

[0021] Further, the radius parameter ε is 10 meters; and / or

[0022] The minimum number of points MinPts is 5.

[0023] Further, compare the first cluster set C and the second cluster set C′ to determine whether a new cluster is generated, including:

[0024] If the Euclidean distance between all the points of a cluster in the second cluster set C′ and the cluster center of the first cluster set C is greater than the set threshold Dthreshold, then determine this cluster as a new cluster;

[0025] Among them, for the cluster Ck, the cluster center μk of the cluster Ck is

[0026]

[0027] where xi is the point in the cluster Ck, and |Ck| is the number of points in the cluster Ck.

[0028] Further, the set threshold Dthreshold is 200 meters.

[0029] The present invention also provides a system for implementing the above - mentioned caregiver going - out detection method based on machine learning. This system includes:

[0030] A location data collection module, configured to collect the location data of the caregiver's current work order using a GPS data acquisition system;

[0031] A preprocessing module, configured to clean the location data, filter out false data, so as to obtain preprocessed location data;

[0032] A decision algorithm module, configured to extract the preprocessed location data and the historical work unit location data of the current work order, and make a decision on the service location of the caregiver this time, so as to determine whether there is an out-of-office behavior; and

[0033] A warning module, configured to issue a warning if it is determined that the caregiver has an out-of-office behavior, so as to remind relevant supervisors;

[0034] Wherein extracting the preprocessed location data and the historical work unit location data of the current work order, and making a decision on the service location of the caregiver this time, so as to determine whether there is an out-of-office behavior includes:

[0035] Extract service point data from the historical work order records of the current work order to form a data set D = {d 1 , d 2 ,..., dn}, where n represents the number of work orders in the data set, and di represents the set of geographical locations served by the current work order in the previous i-th work order;

[0036] Use the DBSCAN algorithm to perform clustering analysis on the data set D to generate a first cluster set C = {C 1 , C 2 ,..., Ck}, where k represents the number of clusters in the cluster set C;

[0037] Obtain the location data dnew of the caregiver's current work order, merge the data set D to form D′, where D′ = D ∪ {dnew}, and use the DBSCAN algorithm to perform clustering analysis on the data set D′ to generate a second cluster set C′ = {C 1 ′, C 2 ′,..., Ck″}, where k′ represents the number of clusters in the cluster set C′;

[0038] Compare the first cluster set C and the second cluster set C′ to determine whether a new cluster is generated; and

[0039] If a new cluster is generated, it is determined that the caregiver has an out-of-office behavior. If no new cluster is generated,

[0040] It is determined that the caregiver does not have an out-of-office behavior.

[0041] Furthermore, the location data, the preprocessed location data, and the historical work unit location data of the current work order are all stored in a database.

[0042] The present invention also provides a computer system, which includes:

[0043] A processor configured to execute machine-readable instructions;

[0044] A graphics card with an artificial intelligence chip, configured to train a machine learning-based caregiver going-out detection method; and

[0045] A memory configured to store machine-readable instructions, the steps of the machine learning-based caregiver going-out detection method when the machine-readable instructions are executed by the processor and / or the graphics card.

[0046] The present invention also provides a computer-readable storage medium, on which machine-readable instructions are stored, the steps of the machine learning-based caregiver going-out detection method when the machine-readable instructions are executed by the processor.

[0047] The technical solution provided by the present invention has the following advantages:

[0048] 1. The machine learning-based caregiver going-out detection method proposed by the present invention, based on machine learning technology, can automatically and intelligently analyze the working positions of caregivers, avoiding the inefficiency and misjudgment of manual monitoring, and effectively solving the problem that the sign-in and clock-in system cannot supervise the going-out behavior of caregivers during the service process.

[0049] 2. The machine learning-based caregiver going-out detection method proposed by the present invention effectively solves the problem that the GPS positioning data is inaccurate and cannot determine the going-out behavior of caregivers by using the clustering algorithm, improving the accuracy and practicability of the caregiver management system.

[0050] 3. The machine learning-based caregiver going-out detection system proposed by the present invention is also applicable to other fields that need to monitor the behaviors of staff, such as medical care, community service, etc., and has broad application prospects and economic value. BRIEF DESCRIPTION OF THE DRAWINGS

[0051] To further clarify the above and other advantages and features of the embodiments of the present invention, more specific descriptions of the embodiments of the present invention will be presented with reference to the accompanying drawings. It can be understood that these drawings only depict typical embodiments of the present invention and will not be considered as limiting its scope. In the drawings, for clarity, the same or corresponding components will be denoted by the same or similar reference numerals.

[0052] Figure 1 A flowchart showing the machine learning-based caregiver going-out detection method according to an embodiment of the present invention;

[0053] Figure 2 A diagram showing clusters formed by historical work position data according to an embodiment of the present invention;

[0054] Figure 3Shows the historical work unit position data of an embodiment of the present invention and the cluster schematic diagram formed by adding the current work unit position data; and

[0055] Figure 4 Shows the schematic diagram of the caregiver going out detection system based on machine learning according to an embodiment of the present invention. Detailed implementation manners

[0056] In the following description, the present invention is described with reference to various embodiments. However, those skilled in the art will recognize that the embodiments can be implemented without one or more specific details or in combination with other alternative and / or additional methods or components. In other cases, well-known structures or operations are not shown or described in detail to avoid obscuring the inventive points of the present invention. Similarly, for the purpose of explanation, specific numbers and configurations are set forth in order to provide a comprehensive understanding of the embodiments of the present invention. However, the present invention is not limited to these specific details.

[0057] In this specification, the reference to "an embodiment" or "the embodiment" means that the specific features, structures, or characteristics described in connection with the embodiment are included in at least one embodiment of the present invention. The phrase "in an embodiment" appearing throughout this specification does not necessarily refer to the same embodiment.

[0058] It should be noted that the embodiments of the present invention describe the method steps in a specific order. However, this is only for the purpose of explaining the specific embodiment and does not limit the order of the steps. On the contrary, in different embodiments of the present invention, the order of the steps can be adjusted according to the actual requirements.

[0059] In the present invention, each module of the system according to the present invention can be implemented using software, hardware, firmware, or a combination thereof. When a module is implemented using software, the functions of the module can be realized through a computer program flow. For example, the module can be implemented by a code segment (such as a code segment in languages like C, C++) stored in a storage device (such as a hard disk, memory, etc.). When the code segment is executed by a processor, the corresponding functions of the module can be realized. When a module is implemented using hardware, the functions of the module can be realized by setting the corresponding hardware structure. For example, the functions of the module can be realized by hardware programming of programmable devices such as field-programmable gate arrays (FPGAs), or by designing an application-specific integrated circuit (ASIC) including multiple electronic devices such as transistors, resistors, and capacitors. When a module is implemented using firmware, the functions of the module can be written in a read-only memory such as EPROM or EEPROM of the device in the form of program code, and when the program code is executed by a processor, the corresponding functions of the module can be realized. Additionally, certain functions of the module may need to be realized by separate hardware or in cooperation with the hardware. For example, the detection function is realized by corresponding sensors (such as proximity sensors, acceleration sensors, gyroscopes, etc.), the signal emission function is realized by corresponding communication devices (such as Bluetooth devices, infrared communication devices, baseband communication devices, Wi-Fi communication devices, etc.), the output function is realized by corresponding output devices (such as displays, speakers, etc.), and so on.

[0060] Traditional caregiver supervision methods are mostly simple GPS positioning and manual spot checks, which have problems such as untimely supervision, inaccurate information, and missed supervision, and cannot effectively prevent caregivers from leaving the elderly's residence during the service process, resulting in potential safety hazards. The present invention provides a method for detecting a caregiver's going out based on machine learning, which uses a machine learning algorithm to solve the problem of inaccurate GPS positions. The algorithm clusters historical multiple inaccurate GPS data, so that the inaccurate data can obtain trustworthy accurate data through aggregation.

[0061] Figure 1 The flowchart of the method for detecting a caregiver's going out based on machine learning according to an embodiment of the present invention is shown. The following combines Figure 1 , and the method for detecting a caregiver's going out based on machine learning proposed by the present invention will be described. In an embodiment of the present invention, the method for detecting a caregiver's going out based on machine learning can be executed by a computer. The method for detecting a caregiver's going out based on machine learning includes the following steps:

[0062] First, use a GPS data acquisition system to collect the location data of the current work order of the caregiver. When the caregiver arrives at the elderly person's home and signs in, the communication device carried by the caregiver, such as a mobile phone, regularly uploads the caregiver's location information to the GPS data acquisition system, for example, once every minute.

[0063] Next, clean the location data to filter out false data in order to obtain preprocessed location data. The false data is mainly data for false clock-in.

[0064] Next, extract the preprocessed location data and the historical work order location data of the current work order, and make a decision on the service location of the caregiver this time to determine whether there is an out-of-office behavior. This step is the decision algorithm. The decision algorithm includes the following 5 steps:

[0065] 1. Extract service point location data from the historical work order records of the current work order to form a data set D = {d 1 , d 2 ,..., dn}, where n represents the number of work orders in the data set, and di represents the set of geographical locations served by the current work order in the first i-th work order.

[0066] 2. Use the DBSCAN algorithm to perform clustering analysis on the data set D to generate a first cluster set C = {C 1 , C 2 ,..., Ck}, where k represents the number of clusters in the cluster set C. Each cluster represents a group of service point locations that are relatively close geographically. The steps to generate the first cluster set C include:

[0067] Initialize the radius parameter ε and the minimum number of points MinPts. The radius parameter ε is used to define the maximum distance of the density neighborhood, that is, the search radius around a point location. In an embodiment of the present invention, the radius parameter ε is 10 meters, that is, if the distance between two GPS point locations in the graph is less than 10 meters, then these two point locations are considered to belong to the same cluster. The minimum number of points MinPts is used to define the minimum number of points included in a cluster. If the number of points in a cluster is less than this value, it is not considered a cluster. In an embodiment of the present invention, the minimum number of points MinPts is 5, that is, a cluster contains at least 5 GPS point locations.

[0068] Determine the core point locations. For each point location in the data set D, if the number of neighborhood point locations within the radius parameter ε is greater than or equal to the minimum number of points MinPts, then determine this point location as a core point location; and

[0069] For each core point, all the points within its ε-radius range are added to the core point to form a cluster. If the neighborhoods of two core points overlap, the clusters where the two core points are located are merged into one cluster. Finally, the dataset is divided into several clusters, and each cluster contains several density-connected points. The points that do not belong to any cluster are marked as noise points, and the noise points are reserved separately.

[0070] 3. Obtain the location data dnew of the current work order of the caregiver, and merge the dataset D to form D′, where D′ = D ∪ {dnew}. Use the DBSCAN algorithm to perform clustering analysis on the dataset D′ to generate the second cluster set C′ = {C 1 ′, C 2 ′,..., Ck″}, where k′ represents the number of clusters in the cluster set C′.

[0071] 4. Compare the first cluster set C and the second cluster set C′ to determine whether a new cluster is generated. If the Euclidean distance between all the points in a cluster in the second cluster set C′ and the cluster center of one of the clusters in the first cluster set C is greater than the set threshold Dthreshold, then determine that this cluster is a new cluster. In an embodiment of the present invention, the set threshold Dthreshold is 200 meters.

[0072] Among them, for the cluster Ck, the cluster center μk of the cluster Ck is

[0073]

[0074] where xi is the point in the cluster Ck, and |Ck| is the number of points in the cluster Ck.

[0075] Further, the set threshold is 200 meters.

[0076] 5. If a new cluster is generated, it is determined that the caregiver has an out-of-office behavior. If no new cluster is generated, it is determined that the caregiver does not have an out-of-office behavior. That is, generating a new cluster can be understood as the caregiver has moved at a location more than 200 meters away from the elderly's home during the current work order, so it is determined that he has an out-of-office behavior.

[0077] Finally, if it is determined that the caregiver has an out-of-office behavior, a warning is issued to remind the relevant supervisors.

[0078] The effect of the fault diagnosis method provided by the present invention can be further illustrated by the following experimental results.

[0079] Figure 2 Shows a schematic diagram of the clusters formed by the historical work order location data in an embodiment of the present invention. Figure 3Shows a cluster schematic diagram formed by historical work unit location data of an embodiment of the present invention and adding current work unit location data. The numerical values behind the clusters have no specific meaning. c-0 refers to the 0th cluster, and c-1 refers to the first cluster. Figure 2 The c-0 in Figure 3 has nothing to do with the c-0 in , only the cluster numbers in the two figures have changed. Figure 2 There are two clusters in , c-0 and c-1. Figure 4 There are 4 clusters in , c-0, c-1, c-2, and c-3. Among them, c-x represents the noise points that cannot be classified into any cluster. The percentage behind the cluster in the figure represents the proportion of the number of all points in the cluster to the total number of points. For example, Figure 2 The 96.7% of c-0 in means that the number of all points in the cluster accounts for 96.7% of the total number of points. It should be noted that setting the threshold Dthreshold to 200 meters means that in the same cluster, there is no point whose distance from other points in the cluster exceeds 200 meters. For example, during the algorithm calculation, assuming that 10 points have been classified into the c-0 cluster, and at this time the 11th point is obtained. If the distance of this point from the 10 points in c-0 exceeds 200 meters, the 11th point will not be classified into the c-0 cluster.

[0080] The caregiver going-out detection method based on machine learning proposed by the present invention, based on machine learning technology, can automatically and intelligently analyze the work locations of caregivers, avoiding the inefficiency and misjudgment of manual monitoring, and effectively solving the problem that the sign-in and clock-in system cannot supervise the going-out behavior of caregivers during the service process; using the clustering algorithm effectively solves the problem that the GPS positioning data is inaccurate and leads to the inability to determine the going-out behavior of caregivers, improving the accuracy and practicality of the caregiver management system.

[0081] In an embodiment of the present invention, the present invention also provides a system for the above-mentioned caregiver going-out detection method based on machine learning. Figure 4 Shows a schematic diagram of a caregiver going-out detection system based on machine learning according to an embodiment of the present invention. As Figure 4 shown, the system includes the following modules:

[0082] A location data collection module, configured to collect the location data of the caregiver's current work order using a GPS data acquisition system;

[0083] A preprocessing module, configured to clean the location data and filter out false data to obtain preprocessed location data;

[0084] A decision algorithm module, configured to extract the preprocessed location data and the historical work unit location data of the current work order, and make a decision on the service location of the caregiver this time to determine whether there is a going-out behavior; and

[0085] A warning module, configured to issue a warning to remind relevant supervisors if it is determined that the caregiver has an out - of - office behavior;

[0086] Among them, extracting the pre - processed location data and the historical work order location data of the current work order, and making a decision on the service location of the caregiver this time to determine whether there is an out - of - office behavior includes:

[0087] Extracting service point data from the historical work order records of the current work order to form a data set D = {d 1 ,d 2 ,...,dn}, where n represents the number of work orders in the data set, and di represents the set of geographical locations served by the current work order in the first i work orders;

[0088] Using the DBSCAN algorithm to perform clustering analysis on the data set D to generate a first cluster set C = {C 1 ,C 2 ,...,Ck}, where k represents the number of clusters in the cluster set C;

[0089] Obtaining the location data dnew of the caregiver's current work order, merging the data set D to form D′, where D′ = D ∪ {dnew}, using the DBSCAN algorithm to perform clustering analysis on the data set D′ to generate a second cluster set C′ = {C 1 ′,C 2 ′,...,Ck″}, where k′ represents the number of clusters in the cluster set C′;

[0090] Comparing the first cluster set C and the second cluster set C′ to determine whether a new cluster is generated; and

[0091] If a new cluster is generated, it is determined that the caregiver has an out - of - office behavior. If no new cluster is generated,

[0092] it is determined that the caregiver does not have an out - of - office behavior.

[0093] In an embodiment of the present invention, the location data, the pre - processed location data, and the historical work order location data of the current work order are all stored in a database.

[0094] The caregiver out - of - office detection system based on machine learning proposed by the present invention is also applicable to other fields that need to monitor the behaviors of staff, such as medical care, community service, etc., and has broad application prospects and economic value.

[0095] In an embodiment of the present invention, the present invention further provides a computer system, which includes a processor, a graphics card with an artificial intelligence chip, and a memory. The memory is configured to store machine-readable instructions, the graphics card is configured to train the machine learning-based caregiver going-out detection method, and the processor is configured to execute the machine-readable instructions. When the processor and / or the graphics card execute the machine-readable instructions, the following processing steps are implemented: collecting the location data of the caregiver's current work order using a GPS data collection system; cleaning the location data to filter out false data to obtain preprocessed location data; extracting the preprocessed location data and the historical work order location data of the current work order, and making a decision on the service location of the caregiver this time to determine whether there is a going-out behavior; and if it is determined that the caregiver has a going-out behavior, issuing a warning to remind relevant supervisors.

[0096] The graphics card may preferably be a graphics card with a GPU computing power higher than model 5.0. Since the amount of data to be trained is large, providing a graphics card configuration can significantly improve the training speed.

[0097] The memory includes various media that can store machine-readable instructions, such as USB flash drives, read-only memories (ROMs), random access memories (RAMs), external hard drives, magnetic disks, or optical discs.

[0098] It can be understood that in addition to the memory and the processor described above, the above computer system further includes other software and hardware components not listed in this specification. Specifically, it can be determined according to the model of the specific data processing device in different application scenarios, and this specification will not list and elaborate one by one.

[0099] In an embodiment of the present invention, the present invention further provides a computer-readable storage medium, on which machine-readable instructions are stored. When the machine-readable instructions are executed by a processor, the following processing steps are implemented: collecting the location data of the caregiver's current work order using a GPS data collection system; cleaning the location data to filter out false data to obtain preprocessed location data; extracting the preprocessed location data and the historical work order location data of the current work order, and making a decision on the service location of the caregiver this time to determine whether there is a going-out behavior; and if it is determined that the caregiver has a going-out behavior, issuing a warning to remind relevant supervisors.

[0100] Although the embodiments of the present invention have been described above, it should be understood that they are presented by way of example only and not as a limitation. It will be apparent to those skilled in the relevant art that various combinations, variations, and changes can be made thereto without departing from the spirit and scope of the present invention. Therefore, the breadth and scope of the present invention disclosed herein should not be limited by the exemplary embodiments disclosed above, but should be defined in accordance with the technical solutions of the present invention and their equivalents.

Claims

1. A method for detecting caregivers going out based on machine learning, characterized in that: The steps include: Use GPS data collection system to collect the location data of the caregiver's current work order; Cleaning the location data and filtering out false data to obtain preprocessed location data; Extracting the pre-processed location data and the historical work order location data of the current work order, and making a decision on the service location of the caregiver this time to determine whether there is an out-of-town behavior; as well as If it is determined that the caregiver has gone out, a warning will be issued to alert the relevant supervisory personnel; The pre-processed location data and the historical work order location data of the current work order are extracted, and a decision is made on the service location of the caregiver to determine whether there is an out-of-town behavior, including: Extract the service point data from the historical work order records of the current work order to form a data set D = {d1, d2, ..., d n }, where n represents the number of work orders in the dataset, and d i Represents the set of geographical locations served by the current work order in the previous i-th work order; Use DBSCAN algorithm to perform cluster analysis on data set D and generate the first cluster set C = {C1, C2, ..., C k }, where k represents the number of clusters in the cluster set C; Get the location data of the caregiver's current work order new , merge the dataset D to form D′, where D′=D∪{d new }, use the DBSCAN algorithm to perform cluster analysis on the data set D′ and generate the second cluster set C′={C′1,C′2,...,C′ k′ }, where k′ represents the number of clusters in the cluster set C′; Compare the first cluster set C with the second cluster set C′ to determine whether to generate a new cluster; and If a new cluster is generated, it is determined that the caregiver has gone out. If no new cluster is generated, it is determined that the caregiver has not gone out.

2. The machine learning-based caregiver outing detection method according to claim 1 is characterized in that: Use DBSCAN algorithm to perform cluster analysis on data set D and generate the first cluster set C = {C1, C2, ..., C k }include: Initialize the radius parameter ε and the minimum number of points MinPts, where the radius parameter ε is used to define the maximum distance of the density neighborhood, and the minimum number of points MinPts is used to define the minimum number of points contained in a cluster; Determine the core point. If the number of neighborhood points of a point in the data set D within the radius parameter ε is greater than or equal to the minimum number of points MinPts, then the point is determined to be a core point; and For each core point, all points within its ε radius plus the core point are taken as a cluster. If the neighborhoods of two core points overlap, the clusters where the two core points are located are merged into one cluster, and the points that do not belong to any cluster are marked as noise points and retained separately.

3. The method for detecting caregivers going out based on machine learning according to claim 2, characterized in that: The radius parameter ε is 10 meters; and / or The minimum number of points MinPts is 5.

4. The method for detecting caregivers going out based on machine learning according to claim 1, characterized in that: Comparing the first cluster set C with the second cluster set C′ to determine whether to generate a new cluster includes: If the Euclidean distance between all points in a cluster of the second cluster set C′ and the cluster center of the first cluster set C is greater than the set threshold D threshold , then the cluster is determined to be a new cluster; Among them, for cluster C k , cluster C k The cluster center μ k for, Among them, x i Cluster C k The point in |C k | is cluster C k The number of midpoints.

5. The method for detecting caregiver absence based on machine learning according to claim 4, characterized in that: The threshold value D threshold is 200 meters.

6. A system for implementing the machine learning-based caregiver out-of-office detection method according to any one of claims 1 to 5, characterized in that: include: A location data collection module is configured to collect location data of the caregiver's current work order using a GPS data collection system; A preprocessing module is configured to clean the location data and filter out false data to obtain preprocessed location data; A decision algorithm module is configured to extract the pre-processed location data and the historical work order location data of the current work order, and make a decision on the service location of the caregiver this time to determine whether there is an out-of-town behavior; as well as A warning module is configured to issue a warning if it is determined that the caregiver has gone out of the house to alert relevant supervisors; The pre-processed location data and the historical work order location data of the current work order are extracted, and a decision is made on the service location of the caregiver to determine whether there is an out-of-town behavior, including: Extract the service point data from the historical work order records of the current work order to form a data set D = {d1, d2, ..., d n }, where n represents the number of work orders in the dataset, and d i Represents the set of geographical locations served by the current work order in the previous i-th work order; Use DBSCAN algorithm to perform cluster analysis on data set D and generate the first cluster set C = {C1, C2, ..., C k }, where k represents the number of clusters in the cluster set C; Get the location data of the caregiver's current work order new , merge the dataset D to form D′, where D′=D∪{d new }, use the DBSCAN algorithm to perform cluster analysis on the data set D′ and generate the second cluster set C′={C′1,C′2,...,C′ k′ }, where k′ represents the number of clusters in the cluster set C′; Compare the first cluster set C with the second cluster set C′ to determine whether to generate a new cluster; and If a new cluster is generated, it is determined that the caregiver has gone out. If no new cluster is generated, it is determined that the caregiver has not gone out.

7. The machine learning-based caregiver outgoing detection system according to claim 6, characterized in that: The location data, the pre-processed location data, and the historical work order location data of the current work order are all stored in a database.

8. A computer system, characterized in that: include: a processor configured to execute machine-readable instructions; a graphics card having an artificial intelligence chip configured to train a machine learning-based caregiver absence detection method; as well as A memory configured to store machine-readable instructions, wherein the machine-readable instructions, when executed by a processor and / or a graphics card, perform the steps of the method according to any one of claims 1 to 5.

9. A computer-readable storage medium, characterized in that: Machine-readable instructions are stored thereon, and when the machine-readable instructions are executed by a processor, the steps of the method according to any one of claims 1 to 5 are performed.

Images