JP2021182242A - Estimation model construction device - Google Patents

Abstract

To obtain effective information that contributes to constructing an estimation model for estimating a user state without using domain knowledge pertaining to locations.SOLUTION: An estimation model construction device 10 comprises: a sensor data acquisition unit 11 acquiring, from one or more sensors that measure physical elements in one or more users, sensor data for each spatial unit of the users; a reference value calculation unit 12 for calculating, with regard to the acquired sensor data for each spatial unit of a plurality of users for reference value calculation, the reference value of sensor data for each spatial unit on the basis of a prescribed method; and an abnormality degree calculation unit 13 for calculating, with regard to the acquired sensor data for each spatial unit of target users for estimation model construction, the target value of sensor data for each spatial unit on the basis of a prescribed method, and calculating the abnormality degree of the sensor data for each spatial unit of the target users on the basis of the target value of sensor data for each spatial unit and the reference value of sensor data for each spatial unit.SELECTED DRAWING: Figure 1

Description

The present disclosure relates to an estimation model construction device that constructs an estimation model for user state estimation.

が知られており、目的変数ｙと説明変数ｘのペアを大規模に与えて推定モデルｆを学習し構築する（即ち、目的変数ｙと説明変数ｘの関係性を習得する）技術が知られている（下記の特許文献１参照）。このような従来技術では、目的変数ｙを推定するために必要な説明変数ｘ（具体的な値ｖ_ｉとして何を与えるか）に関する識者の知見（以下「ドメインの知識」と呼ぶ）を用いることを前提としている。 Conventionally, the basic formula of the following estimation model

Is known, and a technique for learning and constructing an estimation model f by giving a pair of an objective variable y and an explanatory variable x on a large scale (that is, learning the relationship between the objective variable y and the explanatory variable x) is known. (See Patent Document 1 below). In such prior art, the use of the explanatory variables necessary to estimate the response variable y x (specific value v _i What given as either) regarding experts findings (hereinafter referred to as "domain knowledge") Is assumed.

On the other hand, research results are known that the place where a person is and the condition of the person are related, and as a result, the place where the person is located is considered to affect the behavior of the person. For example, it is considered that the state of the target person for each place can be estimated from the sensor data (that is, the data obtained by detecting the change in the state and behavior of the target person) of the geohash unit).

Japanese Unexamined Patent Publication No. 2016-106689

However, in order to estimate the state of the target person from the sensor data for each place as described above, the knowledge of the domain regarding the place is important, but the knowledge of the domain regarding the place here is the state of the place and the person. And it is needed for each of the various patterns of human behavior.

However, since the knowledge of the domain regarding the place cannot be easily acquired by anyone, it is difficult to learn and construct the estimation model f in the situation where there is no knowledge of the domain regarding the place. Therefore, it has been long-awaited to obtain useful information that contributes to constructing an estimation model for user state estimation without using domain knowledge about the location.

It is an object of the present disclosure to obtain useful information that contributes to constructing an estimation model for user state estimation without using domain knowledge about a place in order to solve the above-mentioned problems.

The applicant stated that the location of the person affects the person's condition and behavior, and that the general knowledge of the domain regarding the location is "when there is a change in the user's condition, a sensor that measures physical factors in the user. , When the sensor data for each location (that is, for each spatial unit (for example, geohash unit) changes) ”, and instead of defining explanatory variables, even if you do not have knowledge of the domain regarding the location, By using the degree of abnormality of the sensor data for each location (space unit) described above, "the degree of abnormality of the sensor data for each space unit of the user" is useful information that contributes to constructing an estimation model for estimating the user state. Invented to obtain.

The estimation model construction device according to the present disclosure includes a sensor data acquisition unit that acquires sensor data for each spatial unit of the user from one or more sensors that measure physical elements in one or more users, and the sensor data. A reference value calculation unit that calculates the reference value of the sensor data for each space unit based on a predetermined method for the sensor data for each space unit of multiple users for the reference value calculation acquired by the acquisition unit. With respect to the sensor data for each space unit of the target user for constructing the estimation model acquired by the sensor data acquisition unit, the target value of the sensor data for each space unit is calculated and obtained based on the predetermined method. Abnormality of the sensor data for each space unit of the target user based on the target value of the sensor data for each space unit obtained and the reference value of the sensor data for each space unit obtained by the calculation by the reference value calculation unit. It is provided with an abnormality degree calculation unit for calculating the degree.

In the above estimation model construction device, when the sensor data acquisition unit acquires sensor data for each spatial unit of the plurality of users from a plurality of sensors that measure physical elements in the plurality of users for reference value calculation, the reference value is obtained. The calculation unit calculates the reference value of the sensor data for each space unit based on a predetermined method for the acquired sensor data for each space unit of a plurality of users. Then, when the sensor data acquisition unit acquires the sensor data for each spatial unit of the target user from the sensor that measures the physical element of the target user for constructing the estimation model, the abnormality degree calculation unit acquires the acquired target. For the sensor data for each space unit of the user, the reference value of the sensor data for each space unit is calculated based on a predetermined method, and the target value of the sensor data for each space unit obtained and the above-mentioned calculation are obtained. The degree of abnormality of the sensor data for each space unit of the target user is calculated based on the reference value of the sensor data for each space unit. This is useful information that contributes to building an estimation model for user state estimation by using the degree of anomaly of sensor data for each spatial unit instead of defining explanatory variables, even if there is no knowledge of the domain regarding the location. As a result, the degree of abnormality of the sensor data for each spatial unit of the target user can be obtained.

According to the present disclosure, it is possible to obtain useful information that contributes to constructing an estimation model for user state estimation without using domain knowledge about the location.

It is a functional block diagram which shows the structure of the estimation model construction apparatus. (A) is a flow chart which shows the calculation process of the reference value of the sensor data for each space unit, and (b) is a flow chart which shows the construction process of the user state estimation model based on the degree of anomaly. (A) is a table exemplifying the acquired sensor data of a certain user, and (b) is a table exemplifying the position information of the user. (A) is a table exemplifying the average / variance of the sensor data for each space unit obtained by calculation, and (b) is a diagram showing the magnitude of the average of the sensor data for each space unit two-dimensionally. be. (A) is a table exemplifying the acquired sensor data of the target user, and (b) is a table exemplifying the position information of the target user. It is a table exemplifying the average / variance of the sensor data of the target user for each space unit obtained by calculation. (A) is a diagram two-dimensionally showing the magnitude of the average of the acceleration sensor data for each space unit for the target user, and (b) is the magnitude of the average of the gyro sensor data for each space unit for the target user. Is a diagram two-dimensionally showing, (c) is a diagram showing the screen on / off status for each spatial unit for the target user two-dimensionally, and (d) is a diagram for each spatial unit for the target user. It is a figure which shows the usage situation of the application A two-dimensionally. It is a figure for demonstrating the calculation process of the degree of abnormality for each space unit. It is a figure for demonstrating the process which aggregates the abnormality degree of each sensor on a day-by-day basis. It is a table which shows the example of the acquired user state data. It is a table which shows the example of the abnormality degree of each sensor of the aggregated day. It is a figure which shows the hardware configuration example of the estimation model construction apparatus.

Hereinafter, an embodiment of the estimation model construction device according to the present disclosure will be described with reference to the drawings.

As shown in FIG. 1, the estimation model construction device 10 includes a sensor data acquisition unit 11, a reference value calculation unit 12, an abnormality degree calculation unit 13, a user state data acquisition unit 14, and an estimation model construction unit 15. The functions and operations of each part will be described below.

The sensor data acquisition unit 11 is a functional unit that acquires sensor data for each spatial unit of a user from one or more sensors that measure physical elements in one or more users. Although the details will be described later, in the calculation process of the reference value of the sensor data for each space unit (that is, the value used for comparison with the target value in the abnormality degree calculation described later), the sensor data acquisition unit 11 calculates the reference value. Sensor data for each spatial unit of the plurality of users is acquired from a plurality of sensors that measure physical elements in the plurality of users. Further, in the process of constructing the user state estimation model based on the degree of abnormality, the sensor data acquisition unit 11 obtains sensor data for each spatial unit of the target user from a sensor that measures a physical element of the target user for constructing the estimation model. (That is, the target value in the abnormality degree calculation described later) is acquired.

The reference value calculation unit 12 is a functional unit that calculates the reference value of the sensor data for each space unit based on a predetermined method for the sensor data for each space unit of a plurality of users for the reference value calculation. A specific example of this calculation method will be described later.

The anomaly degree calculation unit 13 is the target of the sensor data for each space unit based on a predetermined method for the sensor data for each space unit of the target user for constructing the estimation model acquired by the sensor data acquisition unit 11. Based on the value calculated and the target value of the sensor data for each space unit obtained and the reference value of the sensor data for each space unit obtained by the calculation by the reference value calculation unit 12, each space unit of the target user It is a functional part that calculates the degree of abnormality of the sensor data of. Further, the abnormality degree calculation unit 13 determines the abnormality degree of the sensor data for each space unit of the target user obtained by the calculation in a predetermined time width unit (hereinafter, "one day unit" as an example) for each sensor. Aggregate as. A specific example of the calculation method by the abnormality degree calculation unit 13 will be described later.

The user status data acquisition unit 14 is a functional unit that acquires user status data representing the status of the target user in the predetermined time width unit (daily unit). As an example of user state data, a stress index is mentioned, and LFHF (Low Frequency High Frequency), which represents the degree (balance) of tension between sympathetic nerve and parasympathetic nerve, which is a general stress index, is generally known. Depending on the situation, it may be acquired as user status data, or two values of whether the obtained LFHF is "higher" or "lower" than a predetermined reference value may be used.

The estimation model building unit 15 uses the degree of abnormality for each sensor in a predetermined time width unit (daily unit) aggregated by the abnormality degree calculation unit 13 as an explanatory variable, and the predetermined time acquired by the user state data acquisition unit 14. It is a functional unit that builds an estimation model that uses user state data in width units (daily units) as the objective variable.

(Process executed in the estimation model building apparatus 10)
Hereinafter, an example of the processing executed by the estimation model building apparatus 10 will be outlined with reference to FIGS. 2 (a) and 2 (b). The process of FIG. 2A is a process for calculating the reference value of the sensor data for each space unit in advance, and the process of FIG. 2B constructs a user state estimation model based on the degree of abnormality of the target user. It is a process for. Since the reference value obtained in the process of FIG. 2 (a) is used in the process of FIG. 2 (b), the process of FIG. 2 (a) is a premise process of the process of FIG. 2 (b).

In the process of FIG. 2A, the sensor data acquisition unit 11 acquires sensor data for each spatial unit of the plurality of users from a plurality of sensors that measure physical elements in the plurality of users for reference value calculation (). Step S1). As a result, for example, acceleration sensor data, gyro sensor data, screen on / off state data, and application startup data for a certain user as shown in FIG. 3A are acquired, and are also shown in FIG. 3B. As such time-series position information of the user, information such as the geohash in which the user stayed, the latitude / longitude of the geohash, and the time when the stay started is acquired.

Next, the reference value calculation unit 12 calculates the average and variance of the sensor data for each space unit (here, geohash) of the plurality of users acquired in step S1 as the reference value of the sensor data for each space unit (here). Step S2). Specifically, from the time-series position information of a user shown in FIG. 3 (b), the user stays in the geohash 123456ab from 0:00:10 to 0:13:50. Therefore, the reference value calculation unit 12 determines that the user's accelerometer data is in the time zone of 13 minutes and 40 seconds while staying in the geohash 123456ab (that is, the time zone from 0:00:10 to 0:13:50). ), The mean and variance of the square root (the magnitude of the vector) of the sum of squares of the accelerometer data x, y, z are calculated. If the user stayed in the geohash 123456ab in other time zones, the accelerometer data x, y, z in the other time zone was also added, and all the time spent in the geohash 123456ab. The accelerometer data x, y, z in the band are grouped together to calculate the mean and variance. The reference value calculation unit 12 also performs the calculation of the mean and variance of the accelerometer data in such a geohash 123456ab for the other geohashes that have stayed (here, the geohash 123456a9). In addition, the calculation of the mean and variance of the acceleration sensor data as described above is also executed for other sensor data. For example, the average and variance of the screen on time may be calculated from the screen on / off state data, and the average and variance of the usage time of a specific app A may be calculated from the data related to the application startup. .. As a result, the average and variance of the various sensor data illustrated in FIG. 4A can be obtained for each spatial unit (here, geohash). In FIG. 4B, as an example, whether the “average” of the accelerometer data is larger or smaller than a predetermined standard value is shown two-dimensionally for each geohash, and diagonal hatching is shown. The geohash applied indicates that the average of the accelerometer data is large, and the geohash with vertical hatching indicates that the average of the accelerometer data is small. In this way, the average magnitude of the acceleration sensor data can be grasped for each geohash. In step S2, the average / variance for each geohash may be calculated for each user, or the average / variance for each geohash may be calculated collectively for all users.

By the process of FIG. 2A as described above, reference values (average and variance) of various sensor data for each spatial unit (geohash) can be obtained.

Next, in the process of FIG. 2B, the sensor data acquisition unit 11 changes from the sensor that measures the physical element of the target user for constructing the estimation model to the sensor for each spatial unit (geohash) of the target user. Acquire data (step S11). As a result, for example, acceleration sensor data, gyro sensor data, screen on / off state data, and application startup data for the target user as shown in FIG. 5A are acquired, and are also shown in FIG. 5B. As time-series position information of the target user, information such as the geohash in which the target user stayed, the latitude / longitude of the geohash, and the time when the stay started is acquired.

Next, the abnormality degree calculation unit 13 calculates the average and variance of the sensor data for each space unit (geohash) of the target user acquired in step S11 as the target value of the sensor data for each space unit (step S12). ). Specifically, from the time-series position information of the target user shown in FIG. 5 (b), the target user is in the time zone from 9:00:00 to 12:30:10 and from 14:40:20 to 17 :. Since it is known that the user was staying in the geohash 123456a9 until 15:00, as an example, for the acceleration sensor data, the anomaly degree calculation unit 13 is the target in the above two time zones in which the geohash 123456a9 was staying. Calculate the average and variance of the square root (the magnitude of the vector) of the sum of squares of the user's accelerometer data x, y, z. The anomaly degree calculation unit 13 also performs the calculation of the mean and variance of the acceleration sensor data in such a geohash 123456a9 for the other geohash (here, the geohash 123456ba) that has stayed. In addition, the calculation of the mean and variance of the acceleration sensor data as described above is also executed for other sensor data. Similar to step S2 in FIG. 2A described above, the average and variance of the screen on time may be calculated from the screen on / off state data, and the data related to the application startup may be used to calculate the average and variance of the screen on time. You may calculate the mean and variance of the usage time. As a result, the mean and variance of the various sensor data illustrated in FIG. 6 can be obtained for each spatial unit (here, geohash). FIG. 7A shows, as an example, whether the “average” of the acceleration sensor data of the target user is larger or smaller than a predetermined standard value in two dimensions for each geohash, and is oblique. The hatched geohash indicates that the average of the accelerometer data is large, and the vertical hatched geohash indicates that the average of the accelerometer data is small, and the geohash is not hatched. The hash indicates that there is no accelerometer data because the target user is not staying. Similarly, FIG. 7 (b) shows the magnitude of the "average" of the gyro sensor data for each geohash of the target user, and FIG. 7 (c) shows the "average" of the screen on time for each geohash of the target user. The magnitude is shown in FIG. 7D, and the magnitude of the "average" of the usage time of the application A for each geohash of the target user is shown two-dimensionally. In step S12, the average / variance for each geohash may be calculated for each target user, or the average / variance for each geohash may be calculated collectively for all the target users.

Next, the abnormality degree calculation unit 13 calculates the abnormality degree of the sensor data for each spatial unit (geohash) of the target user from the target value obtained in step S12 and the reference value obtained in step S2 (step). S13). Any algorithm may be adopted for the calculation of the degree of abnormality. Here, as an example, the anomaly degree calculation unit 13 calculates the sum of the average difference (absolute value) and the variance difference (absolute value) in each geohash as the anomaly degree. An example of calculation of the degree of abnormality of the acceleration sensor data will be described with reference to FIG. Focusing on the "acceleration average" and "acceleration variance" among the "reference values" and "target values" related to the various sensor data shown in FIG. 8, the geohash 123456ab has no stay history of the target user and is the "target value". Since there is no "", the degree of abnormality cannot be calculated, and it becomes "no value".
For Geohash 123456a9, | 10.21−9.81 | ＋ | 0.43−0.02 | ＝ 0.81
For Geohash 123456ba, | 9.81-9.91 | ＋ | 0.06−0.03 | ＝ 0.13
By the calculation, each abnormality degree can be obtained. Similarly, the degree of abnormality is calculated for another sensor data, and the degree of abnormality for each sensor data for each geohash as shown in the lower table of FIG. 8 is obtained.

By the processing up to this point, even if there is no knowledge of the domain regarding the location, the spatial unit (geo) of the target user can be used as effective information that contributes to the construction of an estimation model for estimating the user state based on the degree of abnormality of the sensor data in step S15 described later. It is possible to obtain the degree of abnormality for each sensor data (in hash units).

Then, as shown in FIG. 9, the abnormality degree calculation unit 13 determines the abnormality degree of the sensor data for each geohash of the target user obtained by the calculation in a predetermined time width unit (here, one day unit) for each sensor. Aggregate as the degree of abnormality. Arbitrary algorithm may be adopted for the aggregation here. For example, the "maximum value" of the abnormality degree of each sensor may be the aggregation result, or the "average value" of the abnormality degree of each sensor may be the aggregation result. However, the "minimum value" of the degree of abnormality of each sensor may be used as the aggregation result. In the example of FIG. 9, the "maximum value" of the abnormality degree of each sensor is used as the aggregation result, and the aggregation result of the abnormality degree of each sensor on a certain day (January 1, 2020) is obtained.

Returning to FIG. 2B, in the next step S14, the user state data acquisition unit 14 determines the stress index LFHF as the user state data representing the state of the target user by a method generally known. Obtained in time width units (daily units), and obtain two values, that is, whether the obtained LFHF is "higher" or "lower" than the predetermined reference value. As a result, for example, as shown in FIG. 10, as daily user status data of the target user (user 1), data indicating whether the stress index LFHF is “higher” or “lower” than the reference value is acquired. NS. It is not essential that the process of step S14 is executed after the process of steps S11 to S13, and the process may be executed in parallel with the process of steps S11 to S13.

In the next step S15, as shown in FIG. 11, the estimation model building unit 15 determines the degree of abnormality x (here, x = (here, x = (here, x =)) for each sensor in a predetermined time width unit (daily unit) obtained in step S13. An estimation model f is constructed using v1, v2, ...)) as an explanatory variable and the user state data y (here, the stress index LFHF) in a predetermined time width unit (daily unit) acquired in step S14 as an objective variable. do. The learning for constructing the estimation model may be executed by any algorithm such as machine learning and deep learning.

According to the embodiment of the invention described above, even if there is no knowledge of the domain regarding the location, as effective information that contributes to constructing an estimation model for estimating the user state, the spatial unit (geohash unit) of the target user is used. The degree of abnormality for each sensor data can be obtained, and an estimation model for estimating the user state based on the degree of abnormality of each sensor data can be constructed.

In the above embodiment, the reference value calculation unit 12 and the abnormality degree calculation unit 13 calculate the average and dispersion of the sensor data for each space unit (geohash unit) as the reference value and the target value, and these reference values and the target value are calculated. The example of calculating the degree of anomaly from the target value has been explained, but it is an example that the average and dispersion of the sensor data are used as the basis of the degree of abnormality calculation, and the numerical values other than the average and dispersion of the sensor data are used as the basis of the degree of abnormality calculation. May be good.

[Terms, variants, etc.]
The block diagram used in the description of the above embodiment shows a block of functional units. These functional blocks (components) are realized by any combination of at least one of hardware and software. Further, the method of realizing each functional block is not particularly limited. That is, each functional block may be realized using one physically or logically coupled device, or two or more physically or logically separated devices can be directly or indirectly (eg, for example). , Wired, wireless, etc.) and may be realized using these plurality of devices. The functional block may be realized by combining the software with the one device or the plurality of devices.

Functions include judgment, decision, judgment, calculation, calculation, processing, derivation, investigation, search, confirmation, reception, transmission, output, access, solution, selection, selection, establishment, comparison, assumption, expectation, and assumption. Broadcasting, notifying, communicating, forwarding, configuring, reconfiguring, allocating, mapping, assigning, etc., but limited to these I can't. For example, a functional block (configuration unit) that makes transmission function is called a transmitting unit (transmitting unit) or a transmitter (transmitter). In each case, as described above, the realization method is not particularly limited.

For example, the estimation model building apparatus in one embodiment may function as a computer that performs processing in the present embodiment. FIG. 12 is a diagram showing a hardware configuration example of the estimation model construction device 10. The estimation model building device 10 described above may be physically configured as a computer device including a processor 1001, a memory 1002, a storage 1003, a communication device 1004, an input device 1005, an output device 1006, a bus 1007, and the like.

In the following description, the word "device" can be read as a circuit, a device, a unit, or the like. The hardware configuration of the estimation model construction device 10 may be configured to include one or more of the devices shown in the figure, or may be configured not to include some of the devices.

For each function in the estimation model building apparatus 10, the processor 1001 performs an operation by loading predetermined software (program) on the hardware such as the processor 1001 and the memory 1002, and controls the communication by the communication apparatus 1004. It is realized by controlling at least one of reading and writing of data in the memory 1002 and the storage 1003.

Processor 1001 operates, for example, an operating system to control the entire computer. The processor 1001 may be configured by a central processing unit (CPU) including an interface with a peripheral device, a control device, an arithmetic unit, a register, and the like.

Further, the processor 1001 reads a program (program code), a software module, data, and the like from at least one of the storage 1003 and the communication device 1004 into the memory 1002, and executes various processes according to these. As the program, a program that causes a computer to execute at least a part of the operations described in the above-described embodiment is used. Although it has been described that the various processes described above are executed by one processor 1001, they may be executed simultaneously or sequentially by two or more processors 1001. Processor 1001 may be mounted by one or more chips. The program may be transmitted from the network via a telecommunication line.

The memory 1002 is a computer-readable recording medium, and is composed of at least one such as a ROM (Read Only Memory), an EPROM (Erasable Programmable ROM), an EEPROM (Electrically Erasable Programmable ROM), and a RAM (Random Access Memory). May be done. The memory 1002 may be referred to as a register, a cache, a main memory (main storage device), or the like. The memory 1002 can store a program (program code), a software module, or the like that can be executed to implement the wireless communication method according to the embodiment of the present disclosure.

The storage 1003 is a computer-readable recording medium, for example, an optical disk such as a CD-ROM (Compact Disc ROM), a hard disk drive, a flexible disk, an optical magnetic disk (for example, a compact disk, a digital versatile disk, a Blu-ray). It may consist of at least one (registered trademark) disk), smart card, flash memory (eg, card, stick, key drive), floppy (registered trademark) disk, magnetic strip, and the like. The storage 1003 may be referred to as an auxiliary storage device. The storage medium described above may be, for example, a database, server or other suitable medium containing at least one of the memory 1002 and the storage 1003.

The communication device 1004 is hardware (transmission / reception device) for communicating between computers via at least one of a wired network and a wireless network, and is also referred to as, for example, a network device, a network controller, a network card, a communication module, or the like.

The input device 1005 is an input device (for example, a keyboard, a mouse, a microphone, a switch, a button, a sensor, etc.) that receives an input from the outside. The output device 1006 is an output device (for example, a display, a speaker, an LED lamp, etc.) that outputs to the outside. The input device 1005 and the output device 1006 may have an integrated configuration (for example, a touch panel). Further, each device such as the processor 1001 and the memory 1002 is connected by a bus 1007 for communicating information. The bus 1007 may be configured by using a single bus, or may be configured by using a different bus for each device.

Each aspect / embodiment described in the present disclosure may be used alone, in combination, or may be switched and used according to the execution. Further, the notification of predetermined information (for example, the notification of "being X") is not limited to the explicit one, but is performed implicitly (for example, the notification of the predetermined information is not performed). May be good.

Although the present disclosure has been described in detail above, it is clear to those skilled in the art that the present disclosure is not limited to the embodiments described in the present disclosure. The present disclosure may be implemented as amendments and modifications without departing from the spirit and scope of the present disclosure, which is determined by the description of the scope of claims. Therefore, the description of this disclosure is for purposes of illustration and does not have any limiting meaning to this disclosure.

The order of the processing procedures, sequences, flowcharts, etc. of each aspect / embodiment described in the present disclosure may be changed as long as there is no contradiction. For example, the methods described in the present disclosure present elements of various steps using exemplary order, and are not limited to the particular order presented.

The input / output information and the like may be stored in a specific place (for example, a memory), or may be managed using a management table. Information to be input / output may be overwritten, updated, or added. The output information and the like may be deleted. The input information or the like may be transmitted to another device.

The phrase "based on" as used in this disclosure does not mean "based on" unless otherwise stated. In other words, the statement "based on" means both "based only" and "at least based on".

When "include", "including" and variations thereof are used in the present disclosure, these terms are as inclusive as the term "comprising". Is intended. Moreover, the term "or" used in the present disclosure is intended not to be an exclusive OR.

In the present disclosure, if articles are added by translation, for example a, an and the in English, the disclosure may include the plural nouns following these articles.

In the present disclosure, the term "A and B are different" may mean "A and B are different from each other". The term may mean that "A and B are different from C". Terms such as "separate" and "combined" may be interpreted in the same way as "different".

10 ... estimation model construction device, 11 ... sensor data acquisition unit, 12 ... reference value calculation unit, 13 ... abnormality degree calculation unit, 14 ... user state data acquisition unit, 15 ... estimation model construction unit, 1001 ... processor, 1002 ... memory , 1003 ... Storage, 1004 ... Communication device, 1005 ... Input device, 1006 ... Output device, 1007 ... Bus.

Claims (3)

A sensor data acquisition unit that acquires sensor data for each spatial unit of the user from one or more sensors that measure physical elements in one or more users.
Reference value calculation for calculating the reference value of the sensor data for each space unit based on a predetermined method for the sensor data for each space unit of a plurality of users for calculating the reference value acquired by the sensor data acquisition unit. Department and
With respect to the sensor data for each space unit of the target user for constructing the estimation model acquired by the sensor data acquisition unit, the target value of the sensor data for each space unit is calculated and obtained based on the predetermined method. Abnormality of the sensor data for each space unit of the target user based on the target value of the sensor data for each space unit obtained and the reference value of the sensor data for each space unit obtained by the calculation by the reference value calculation unit. Abnormality calculation unit that calculates the degree, and
Estimated model building device. The abnormality degree calculation unit aggregates the abnormality degree of the sensor data for each space unit of the target user obtained by the calculation as the abnormality degree for each sensor in a predetermined time width unit.
The estimation model construction device is
A user state data acquisition unit that acquires user state data representing the state of the target user in the predetermined time width unit, and a user state data acquisition unit.
The degree of abnormality for each sensor in the predetermined time width unit aggregated by the abnormality degree calculation unit is used as an explanatory variable, and the user state data in the predetermined time width unit acquired by the user state data acquisition unit is used as the objective variable. The estimation model building unit that builds the estimation model,
The estimation model building apparatus according to claim 1, further comprising. The reference value calculation unit calculates the average and variance of the sensor data for each space unit as reference values.
The anomaly degree calculation unit calculates the average and variance of the sensor data for each space unit as target values.
The estimation model building apparatus according to claim 1 or 2.

Images