WO2016063433A1 - 推定結果表示システム、推定結果表示方法および推定結果表示プログラム - Google Patents
推定結果表示システム、推定結果表示方法および推定結果表示プログラム Download PDFInfo
- Publication number
- WO2016063433A1 WO2016063433A1 PCT/JP2015/002297 JP2015002297W WO2016063433A1 WO 2016063433 A1 WO2016063433 A1 WO 2016063433A1 JP 2015002297 W JP2015002297 W JP 2015002297W WO 2016063433 A1 WO2016063433 A1 WO 2016063433A1
- Authority
- WO
- WIPO (PCT)
- Prior art keywords
- estimation result
- learning model
- estimation
- graph
- symbol
- Prior art date
Links
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F9/00—Arrangements for program control, e.g. control units
- G06F9/06—Arrangements for program control, e.g. control units using stored programs, i.e. using an internal store of processing equipment to receive or retain programs
- G06F9/44—Arrangements for executing specific programs
- G06F9/451—Execution arrangements for user interfaces
- G06F9/453—Help systems
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
- G06F16/90—Details of database functions independent of the retrieved data types
- G06F16/901—Indexing; Data structures therefor; Storage structures
- G06F16/9024—Graphs; Linked lists
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F3/00—Input arrangements for transferring data to be processed into a form capable of being handled by the computer; Output arrangements for transferring data from processing unit to output unit, e.g. interface arrangements
- G06F3/01—Input arrangements or combined input and output arrangements for interaction between user and computer
- G06F3/048—Interaction techniques based on graphical user interfaces [GUI]
- G06F3/0481—Interaction techniques based on graphical user interfaces [GUI] based on specific properties of the displayed interaction object or a metaphor-based environment, e.g. interaction with desktop elements like windows or icons, or assisted by a cursor's changing behaviour or appearance
- G06F3/0482—Interaction with lists of selectable items, e.g. menus
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F3/00—Input arrangements for transferring data to be processed into a form capable of being handled by the computer; Output arrangements for transferring data from processing unit to output unit, e.g. interface arrangements
- G06F3/01—Input arrangements or combined input and output arrangements for interaction between user and computer
- G06F3/048—Interaction techniques based on graphical user interfaces [GUI]
- G06F3/0484—Interaction techniques based on graphical user interfaces [GUI] for the control of specific functions or operations, e.g. selecting or manipulating an object, an image or a displayed text element, setting a parameter value or selecting a range
- G06F3/04842—Selection of displayed objects or displayed text elements
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F8/00—Arrangements for software engineering
- G06F8/30—Creation or generation of source code
- G06F8/38—Creation or generation of source code for implementing user interfaces
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N20/00—Machine learning
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T11/00—2D [Two Dimensional] image generation
- G06T11/20—Drawing from basic elements, e.g. lines or circles
- G06T11/206—Drawing of charts or graphs
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2200/00—Indexing scheme for image data processing or generation, in general
- G06T2200/24—Indexing scheme for image data processing or generation, in general involving graphical user interfaces [GUIs]
Definitions
- the present invention relates to an estimation result display system, an estimation result display method, and an estimation result display program for displaying an estimation result derived using a learning model.
- Non-Patent Document 1 describes that a prediction formula is automatically selected from a plurality of prediction formulas, and a prediction value is calculated using the prediction formula.
- Non-Patent Document 1 describes that a graph indicating the calculated predicted value and the transition of the actual value corresponding to the predicted value is displayed. Further, Non-Patent Document 1 describes that a graph indicating the transition of the selected prediction formula is displayed.
- Non-Patent Document 1 shows an example of a display screen that displays a graph indicating a transition of a predicted value and a graph indicating a transition of a selected prediction formula in different areas.
- Non-Patent Document 1 when an estimation result such as a predicted value is derived, a learning model is automatically selected from a plurality of learning models, and the estimation result using the learning model is selected. May be derived.
- the present inventor considered that it is preferable to display the estimation result so that a human can intuitively recognize which learning model is selected when the estimation result is derived.
- the estimation result is so that humans can intuitively recognize the selected learning model with low estimation accuracy.
- the present inventor considered that it is preferable to display.
- the present invention provides an estimation result display system and an estimation result display method capable of displaying an estimation result so that a human can intuitively recognize which learning model is selected when deriving the estimation result. And it aims at providing an estimation result display program.
- the present invention allows a human to intuitively recognize a selected learning model with low estimation accuracy when a learning model with low estimation accuracy is selected when deriving an estimation result.
- Another object of the present invention is to provide an estimation result display system, an estimation result display method, and an estimation result display program capable of displaying an estimation result.
- the estimation result display system is an input means for inputting information in which an estimation result is associated with information representing a learning model used to derive the estimation result, and the estimation result is represented by a symbol. It is a graph, Comprising: The display means which displays the graph which changed the classification of the symbol according to the learning model corresponding to the estimation result, It is characterized by the above-mentioned.
- the estimation result display system includes an input means for inputting information in which an estimation result is associated with information representing a learning model used when the estimation result is derived, and the estimation result as a symbol.
- the estimation result display system includes an input means for inputting information in which an estimation result is associated with information representing a learning model used to derive the estimation result, and a symbol representing the estimation result. Are displayed in a predetermined order, and display means is provided for sorting and displaying areas in the graph according to the learning model corresponding to the estimation result indicated by the symbol.
- the estimation result display system includes an input means for inputting information in which an estimation result and information representing a learning model used for deriving the estimation result are associated, and a transition of the estimation result.
- the estimation result display method accepts input of information in which an estimation result and information representing a learning model used to derive the estimation result are associated, and a graph representing the estimation result as a symbol And the graph which changed the classification of the symbol according to the learning model corresponding to the estimation result is displayed, It is characterized by the above-mentioned.
- the estimation result display method accepts input of information in which an estimation result and information representing a learning model used to derive the estimation result are associated, and a graph representing the estimation result as a symbol When any symbol in the graph receives a selection operation, information representing a learning model corresponding to an estimation result indicated by the symbol that has received the selection operation is displayed.
- the estimation result display method accepts input of information in which an estimation result and information representing a learning model used to derive the estimation result are associated with each other, and a symbol representing the estimation result is set to a predetermined symbol.
- the graphs arranged in order are displayed, and the regions in the graph are divided and displayed according to the learning model corresponding to the estimation result indicated by the symbol.
- the estimation result display method accepts input of information in which an estimation result is associated with information representing a learning model used when the estimation result is derived, and is a line graph representing a transition of the estimation result And the graph which changed the attribute of the line according to the learning model corresponding to an estimation result is displayed, It is characterized by the above-mentioned.
- the estimation result display program is mounted on a computer having an input unit for inputting information in which an estimation result is associated with information representing a learning model used when the estimation result is derived.
- An estimation result display program for causing a computer to execute a display process for displaying a graph representing an estimation result as a symbol and changing a symbol type according to a learning model corresponding to the estimation result. It is characterized by that.
- the estimation result display program is mounted on a computer having an input unit for inputting information in which an estimation result is associated with information representing a learning model used when the estimation result is derived.
- the estimation result display program displays a graph representing the estimation result as a symbol on a computer, and when any symbol in the graph is subjected to a selection operation, the symbol that has received the selection operation indicates A display process for displaying information representing a learning model corresponding to the estimation result is executed.
- the estimation result display program is mounted on a computer having an input unit for inputting information in which an estimation result is associated with information representing a learning model used when the estimation result is derived.
- a display process for displaying is executed.
- the estimation result display program is mounted on a computer having an input unit for inputting information in which an estimation result is associated with information representing a learning model used when the estimation result is derived.
- An estimation result display program for causing a computer to execute a display process for displaying a graph representing a transition of an estimation result, wherein the graph changes a line attribute according to a learning model corresponding to the estimation result. It is characterized by.
- the present invention it is possible to display the estimation result so that a human can intuitively recognize which learning model is selected when the estimation result is derived.
- the present invention when a learning model with low estimation accuracy is selected when deriving an estimation result, a human can intuitively recognize the selected learning model with low estimation accuracy.
- the estimation result can be displayed as possible.
- FIG. 1 is a schematic diagram showing a learning device and an estimator.
- description will be made using a specific example in which the value of an objective variable such as water consumption is estimated (predicted) based on the values of explanatory variables such as temperature, precipitation, and wind speed.
- the learning device 11 generates a plurality of learning models by using learning data in advance.
- the learning model is a model for deriving an estimation result when estimation data is given. In other words, an estimation result is obtained by applying the learning model to the estimation data.
- the learning model is information indicating regularity established between the explanatory variable and the objective variable, which is derived from the learning data, for example.
- the learning model is generated in the form of an estimation formula. In this case, the estimation result is calculated by substituting the estimation data into the explanatory variable of the estimation formula.
- the case where the learning model is in the form of the estimation formula is taken as an example, but the format of the learning model is not necessarily the estimation formula.
- a plurality of learning models generated by the learning device 11 is used by the estimator 12.
- estimation data is input to the estimator 12, and the estimator 12 selects a learning model according to a condition satisfied by the estimation data from a plurality of learning models. Then, the estimator 12 derives an estimation result using the estimation data and the selected learning model.
- the estimator 12 may calculate the estimation result by substituting the input estimation data into the explanatory variable of the estimation formula. .
- a plurality of sets of information associating the estimation results derived by the estimator 12 with the information representing the learning model used when deriving the estimation results are input to the estimation result display system 1 of the present invention.
- Information representing the learning model is identification information of the learning model.
- other information is also associated with the information and input to the estimation result display system 1.
- Each estimation result input to the estimation result display system 1 of the present invention is derived in advance by the estimator 12.
- FIG. 2 is a schematic diagram illustrating an example of a selection model.
- the selection model is a tree-structure model in which a learning model is a leaf node and a condition related to estimation data is defined for nodes other than the leaf node is illustrated.
- each node other than the leaf node has two child nodes.
- the selection model is a tree structure model illustrated in FIG. 2 will be described as an example.
- the format of the selection model is not limited to the tree structure model.
- the selection model is also given to the estimator 12. Assume that the selection model illustrated in FIG. 2 is given to the estimator 12 and that the temperature and precipitation values are input to the estimator 12 as estimation data. Then, the estimator 12 starts from the root node of the selection model and repeats selecting one of the two child nodes depending on whether or not the estimation data satisfies the condition indicated by the node. follow. Then, when the estimator 12 arrives at the leaf node, the estimator 12 selects the learning model indicated by the leaf node. Then, the estimator 12 derives an estimation result using the learning model and the estimation data. In the above example, the case where the temperature and precipitation values described as conditions in the selection model shown in FIG. 2 are input as estimation data is exemplified. However, items that are not described as conditions in the selection model are for estimation. It may be included in the data.
- FIG. 3 is a diagram illustrating a specific example of the estimation data input to the estimator 12.
- FIG. 3 shows a set of estimation data.
- Information corresponding to “row” in FIG. 3 is information corresponding to one estimation data.
- the estimation data includes, for example, a plurality of attributes.
- the information corresponding to the “column” in FIG. 3 is information indicating the attributes constituting the estimation data.
- the estimation data includes an ID (identifier) for identifying the estimation data, a temperature value, a precipitation value, a wind speed value, and information indicating time.
- the set of estimation data is expressed in a table format, but the estimation data is not limited to the format shown in FIG.
- the estimator 12 calculates the estimation result by, for example, substituting the attribute value included in the estimation data into the explanatory variable of the estimation formula.
- FIG. 4 is a diagram illustrating a specific example of estimation result data that is information output from the estimator 12.
- FIG. 4 shows a set of estimation result data.
- Information corresponding to “row” in FIG. 4 is information corresponding to one estimation result data.
- the estimation result data is, for example, information in which an estimated value is associated with information representing a learning model used when the estimated value is derived.
- the estimation result data may include other information.
- the estimation result data includes an identifier for identifying the estimation result data, an identifier for the estimation data that is a basis for calculating the estimation value, time information of the estimation data, and the like. Shows the case.
- the estimation result data may include an attribute value included in the estimation data that is a basis for calculating the estimated value.
- the set of estimation result data is information indicating a series of estimation results that are continuous in time series, for example.
- an example of the operation of the estimator 12 has been described using a specific example.
- the selection model as described above may be referred to as a learning model, and a model for deriving an estimation result, such as an estimation formula, may be referred to as a component.
- a model for example, an estimation formula
- a model for selecting a learning model is referred to as a selection model.
- the selection model itself is a learning result.
- the selection model may be a learning result or manually generated information. Good.
- FIG. FIG. 5 is a block diagram illustrating a configuration example of the estimation result display system according to the first embodiment of this invention.
- the estimation result display system 1 of the present embodiment includes an input unit 2 and a display unit 3.
- the input means 2 receives a plurality of sets of information (that is, estimation result data) in which an estimation result and information representing a learning model used for deriving the estimation result are associated with each other.
- a set of estimation result data is input to the input unit 2.
- the estimation result included in the estimation result data is derived in advance by the estimator 12 (see FIG. 1).
- the learning model used when deriving the estimation result is selected by the estimator 12, for example.
- the estimation result may be referred to as an estimated value.
- the estimation result and information representing the learning model are associated with each other. Therefore, it can be said that the estimation result is also associated with the learning model itself.
- the input unit 2 sends the input information to the display unit 3.
- the input unit 2 is realized by an input device or an input interface for inputting information, for example.
- the display means 3 is a graph that represents the estimation result as a symbol, and displays a graph in which the symbol type is changed according to the learning model corresponding to the estimation result.
- the symbol is also referred to as a marker.
- FIG. 6 is a schematic diagram illustrating an example of a graph displayed by the display unit 3 in the first embodiment.
- the horizontal axis of the graph displayed in the present embodiment is an axis representing the order of estimation results.
- FIG. 6 illustrates a case where the horizontal axis of the graph represents time. In this case, it suffices if information on the corresponding time is added to the estimation result input to the input unit 2.
- a set of estimation results and information representing a learning model that is, estimation result data
- the horizontal axis of the graph may not be an axis representing time, but may be an axis representing the input order of estimation result data, for example.
- the vertical axis of the graph is the axis corresponding to the estimation result.
- the display means 3 displays a graph in which symbols representing the estimation results are arranged in a predetermined order (in this example, the order of times corresponding to the estimation results). More specifically, the display unit 3 calculates each symbol representing the estimation result as a value while arranging the symbols in a predetermined order (in this example, the order of times corresponding to the estimation result) along the horizontal axis. It arrange
- the display unit 3 uses the symbol representing the estimation result as the x coordinate (coordinate in the horizontal axis) as the order of the estimation result (for example, the order represented by the time), and the estimation result as the y coordinate ( It is arranged at a position to be the coordinate in the vertical axis direction).
- the interval between symbols arranged in order is short, it can be recognized as a line graph.
- FIG. 6 illustrates a case where the interval between symbols arranged in order is short and is recognized as a line graph.
- the display means 3 changes the type of symbol representing the estimation result according to the learning model corresponding to the estimation result. That is, the display means 3 represents a symbol representing an estimation result derived using the learning model 1, a symbol representing an estimation result derived using the learning model 2, and an estimation result derived using the learning model 3. Different symbols are arranged in the graph as symbols or the like. The same applies when there are more types of learning models selected when the estimation result is derived.
- the display means 3 may change, for example, the color of the symbol or the shape of the symbol when changing the type of the symbol depending on the type of the learning model.
- the method of changing the symbol type is not particularly limited.
- the display means 3 is realized by a CPU of a computer having a display device.
- the CPU may read the estimation result display program from a program recording medium such as a computer program storage device (not shown in FIG. 5) and operate as the display unit 3 according to the estimation result display program.
- a part of the display means 3 that defines a graph and causes the display device to display the graph is realized by the CPU.
- the part that actually performs display is realized by a display device. This point is the same in each embodiment described later.
- estimation result display system 1 may have a configuration in which two or more physically separated devices are connected by wire or wirelessly. This also applies to each embodiment described later.
- FIG. 7 is a flowchart illustrating an example of processing progress of the first embodiment.
- a plurality of sets of information in which an estimation result and information representing a learning model used in deriving the estimation result are associated with each other are input to the input unit 2.
- the input unit 2 sends the input information to the display unit 3 (step S1).
- the display means 3 is a graph in which the symbols representing the estimation results are arranged so that the respective estimation results calculated as values are the coordinates in the vertical axis while arranging the symbols along the horizontal axis in a predetermined order.
- a graph is displayed in which the type of symbol representing the estimation result is changed according to the learning model corresponding to the estimation result (step S2).
- the display means 3 displays the graph illustrated in FIG.
- the “predetermined order” may be, for example, a time order corresponding to the estimation results, an input order of the estimation results, or another order.
- the display unit 3 changes the type of the symbol representing the estimation result according to the learning model corresponding to the estimation result. Therefore, the observer of the graph can intuitively recognize which learning model was selected when deriving the estimation result represented by the symbol.
- Non-Patent Document 1 a graph representing the predicted value and a graph representing the transition of the prediction formula are displayed side by side.
- an observer of a graph can recognize from a single graph which learning model has been selected when the estimation result represented by the symbol is derived.
- the learning model corresponding to the estimation data is not necessarily the learning model selected by the estimator 12.
- an analyst may manually select a learning model corresponding to the estimation data.
- the estimation result and the information representing the learning model used when deriving the estimation result may be manually associated with each other. The same applies to other embodiments.
- the display means 3 does not necessarily need to change the type of symbol for every learning model.
- the display unit 3 may group learning models, and may change the type of symbol representing an estimation result derived from the learning model for each grouped learning model.
- the display means 3 may change the symbol type representing the estimation result derived from the specific learning model among the plurality of learning models with the type of other symbols. The same applies to other embodiments.
- Embodiment 2 FIG.
- the estimation result display system according to the second embodiment of the present invention can be represented by the block diagram shown in FIG. 5 similarly to the estimation result display system according to the first embodiment. An embodiment will be described. Explanation of matters similar to those in the first embodiment will be omitted as appropriate.
- the estimation result display system 1 of the second embodiment includes an input unit 2 and a display unit 3 (see FIG. 5).
- the estimation result data received by the input unit 2 includes an actual measurement value corresponding to the estimation result.
- the input means 2 includes information (estimation result data) in which an estimation result, information representing a learning model used to derive the estimation result, and an actual measurement value corresponding to the estimation result are associated with each other. ) Are input in multiple sets.
- the input unit 2 sends the input information to the display unit 3.
- the estimation result may be referred to as an estimated value.
- the display means 3 is a graph that represents the estimation result as a symbol, and displays a graph in which the symbol type is changed according to the learning model corresponding to the estimation result.
- the display means 3 represents one estimation result data by one symbol. This is expressed as “the estimation result is represented by a symbol”.
- FIG. 8 is a schematic diagram showing an example of a graph displayed by the display means 3 in the second embodiment.
- the display means 3 displays a scatter diagram in which the estimation results are represented by symbols.
- the scatter diagram displayed in the second embodiment has an axis corresponding to the estimation result and an axis corresponding to the actual measurement value.
- FIG. 8 illustrates a case where the axis corresponding to the estimation result is the horizontal axis and the axis corresponding to the actual measurement value is the vertical axis.
- the display means 3 arranges the symbols representing the estimation results in the scatter diagram so that the estimation results become the coordinates in the horizontal axis direction and the actually measured values corresponding to the estimation results become the coordinates in the vertical axis direction. Further, the type of symbol representing the estimation result is changed according to the learning model corresponding to the estimation result.
- a symbol representing an estimation result derived using the learning model 1 is a circle
- a symbol representing an estimation result derived using the learning model 2 is a square
- an estimation derived using the learning model 3 The case where the symbol showing a result is an equilateral triangle is illustrated. Even when there are more types of learning models to be selected when the estimation result is derived, the display unit 3 changes the type of the symbol according to the learning model.
- FIG. 8 illustrates the case where the display unit 3 changes the symbol shape according to the learning model, but the display unit 3 may change the color of the symbol, for example.
- the method of changing the symbol type is not particularly limited.
- FIG. 9 is a flowchart illustrating an example of processing progress of the second embodiment.
- a plurality of sets of information in which an estimation result, information representing a learning model used when deriving the estimation result, and an actual measurement value are associated are input to the input unit 2.
- the input unit 2 sends the input information to the display unit 3 (step S11).
- the display means 3 is a scatter diagram in which symbols representing estimation results are arranged such that the estimation results are the coordinates in the horizontal axis direction, and the actual measurement values corresponding to the estimation results are the coordinates in the vertical axis direction.
- a scatter diagram in which the type of symbol representing the result is changed according to the learning model corresponding to the estimation result is displayed (step S12). For example, the display means 3 displays the graph illustrated in FIG.
- the display means 3 displays a scatter diagram having a horizontal axis corresponding to the estimation result and a vertical axis corresponding to the actual measurement value, as illustrated in FIG.
- the display means 3 arranges the symbols representing the estimation results in the scatter diagram so that the estimation results are the coordinates in the horizontal axis direction and the actual measurement values corresponding to the estimation results are the coordinates in the vertical axis direction.
- the type of symbol representing the estimation result is changed according to the learning model corresponding to the estimation result.
- the estimation result is x and the actual measurement value is represented by y
- the closer the symbol is to the straight line y x (the straight line indicated by the broken line in FIG. 8), the higher the estimation accuracy of the learning model.
- the estimation accuracy of the learning model selected when the estimation result represented by that symbol is derived is low.
- the estimation result derived by the learning model 2 is represented by a square symbol.
- a human when a learning model with low estimation accuracy is selected when deriving an estimation result, a human can intuitively understand the selected learning model with low estimation accuracy. Can be recognized.
- the observer of the graph (scatter diagram) illustrated in FIG. 8 can intuitively determine the tendency for each learning model at a glance.
- the learning model 2 illustrated in FIG. 8 has a tendency to derive a small estimated value with respect to the actual measurement value
- the learning model 3 has a tendency to derive a large estimated value with respect to the actual measurement value. You can see at a glance from the observer.
- the display means 3 changes the type of the symbol representing the estimation result according to the learning model corresponding to the estimation result. Therefore, the observer of the scatter diagram can intuitively recognize from a single scatter diagram which learning model was selected when deriving the estimation result represented by the symbol in the scatter diagram. .
- the axis corresponding to the actual measurement value may be the horizontal axis
- the axis corresponding to the estimation result may be the vertical axis
- FIG. 5 The estimation result display system according to the third embodiment of the present invention can be represented by the block diagram shown in FIG. 5 similarly to the estimation result display system according to the first embodiment. An embodiment will be described. Explanation of matters similar to those in the first embodiment will be omitted as appropriate.
- the estimation result display system 1 includes an input unit 2 and a display unit 3 (see FIG. 5).
- the estimation result data received by the input unit 2 includes two or more types of attribute values used when deriving the estimation result.
- the input means 2 includes an estimation result, information representing the learning model used to derive the estimation result, and two or more types of attribute values used to derive the estimation result. Are associated with each other (estimation result data).
- the estimation result is derived based on at least two types of attributes.
- the attribute value included in each estimation result data may be different, but the attribute type included in the estimation result data is common to each estimation result data. For example, if the estimation result data includes temperature measurements and precipitation measurements, these measurements may differ for each estimation result data. A measure of quantity shall be included.
- the estimation result may be referred to as a discrimination result.
- the estimator 12 selects the learning model using the measured value of the traffic volume of the vehicle and the measured value of the NOx concentration in the air as the attribute values. Then, the estimator 12 determines whether the traffic and NOx concentration measurement points are “city” or “countryside” based on the measured value of the traffic volume of the vehicle, the measured value of the NOx concentration in the air, and the selected learning model.
- a case where the above is estimated (discriminated) will be described as an example.
- the estimation result may be estimated based on yet another attribute.
- the estimation result data may include attribute values other than the measured value of the traffic volume of the vehicle and the measured value of the NOx concentration in the air.
- the display means 3 is a graph that represents the estimation result as a symbol, and displays a graph in which the symbol type is changed according to the learning model corresponding to the estimation result.
- the display means 3 represents one estimation result data by one symbol. This is expressed as “the estimation result is represented by a symbol”.
- FIG. 10 is a schematic diagram illustrating an example of a graph displayed by the display unit 3 in the third embodiment.
- the display means 3 displays a scatter diagram in which the estimation results are represented by symbols.
- the scatter diagram displayed in the third embodiment shows the axis corresponding to the first attribute (in this example, the traffic volume of the vehicle) used when deriving the estimation result, and when deriving the estimation result.
- an axis corresponding to the second attribute used NOx concentration in air in this example.
- FIG. 10 illustrates the case where the axis corresponding to the traffic volume of the vehicle is the horizontal axis and the axis corresponding to the NOx concentration in the air is the vertical axis.
- the display means 3 arranges the symbol representing the estimation result at a position having two types of attribute values included in the estimation result data corresponding to the estimation result as coordinates.
- the display means 3 uses the symbol representing the estimation result as the measured value of the traffic volume of the vehicle corresponding to the estimation result as x-coordinate (coordinate in the horizontal axis direction), and in the air corresponding to the estimation result.
- the measurement value of NOx is arranged at a position where the y coordinate (ordinate in the vertical axis direction) is used.
- the display means 3 changes the symbol type representing the estimation result according to the estimation result (“city” or “countryside”), and further changes the symbol type according to the learning model corresponding to the estimation result. Also change.
- the display unit 3 has a round symbol when the estimation result is “city” and a cross symbol when the estimation result is “countryside”. Further, when the estimation result represented by the symbol corresponds to the learning model 1, the display unit 3 displays the symbol with a solid line, and the estimation result represented by the symbol corresponds to the learning model 2. The symbol is displayed with a dotted line. That is, in the example shown in FIG. 10, the display unit 3 changes the symbol type according to two criteria: the type of estimation result and the type of learning model corresponding to the estimation result.
- FIG. 11 is a flowchart illustrating an example of processing progress of the third embodiment.
- the input means 2 is associated with an estimation result, information representing the learning model used when deriving the estimation result, and two or more types of attributes used when deriving the estimation result. Multiple sets of information are input.
- the input unit 2 sends the input information to the display unit 3 (step S21).
- the display means 3 is a scatter diagram in which symbols representing estimation results are arranged at positions having coordinates of two types of attribute values corresponding to the estimation results, and the symbol type is changed according to the estimation results. Then, a scatter diagram in which the symbol type is changed in accordance with the learning model corresponding to the estimation result is displayed (step S22). For example, the display unit 3 displays a scatter diagram illustrated in FIG.
- the display unit 3 changes the type of the symbol representing the estimation result according to the learning model corresponding to the estimation result. Therefore, the observer of the scatter diagram can intuitively recognize from a single scatter diagram which learning model was selected when deriving the estimation result represented by the symbol in the scatter diagram. .
- the observer can grasp at a glance the tendency for each learning model from the scatter diagram illustrated in FIG.
- the learning model 1 is selected when the NOx concentration in the air is higher than the traffic volume of the vehicle, and the NOx concentration in the air is lower than the traffic volume of the vehicle.
- the observer can grasp at a glance that the learning model 2 tends to be selected. Further, for example, the observer can grasp at a glance that the learning model 2 tends to be determined as “countryside” as a whole.
- FIG. 12 is a block diagram illustrating a configuration example of the fourth exemplary embodiment of the present invention.
- the estimation result display system 1 according to the fourth embodiment includes an input unit 2, a display unit 3, and a cursor operation unit 7. Note that description of matters similar to those in the first embodiment will be omitted as appropriate.
- the input unit 2 of the fourth embodiment as in the first embodiment, information in which an estimation result and information representing a learning model used to derive the estimation result are associated with each other, Multiple sets are input.
- the input unit 2 sends the input information to the display unit 3.
- the estimation result is calculated as a value will be described as an example.
- Cursor operating means 7 is a device for a graph observer (hereinafter simply referred to as an observer) to operate a cursor in the graph display screen.
- the cursor operation means 7 is a pointing device such as a mouse, a touch pad, a joystick, or a trackball.
- Display means 3 displays a graph representing the estimation result as a symbol.
- FIG. 13 is a schematic diagram illustrating an example of a graph displayed by the display unit 3 in the fourth embodiment.
- FIG. 13A shows a graph in the displayed initial state.
- FIG. 13B shows a graph when a symbol representing an estimation result is pointed and clicked.
- the horizontal axis of the graph is an axis representing the order of estimation results.
- FIG. 13 illustrates a case where the horizontal axis of the graph represents time. In this case, it suffices if information on the corresponding time is added to the estimation result input to the input unit 2.
- a set of an estimation result and information representing a learning model may be input to the input unit 2 in order of time corresponding to the estimation result.
- the vertical axis of the graph is the axis corresponding to the estimation result.
- the display means 3 displays a graph in which symbols representing the estimation results are arranged in a predetermined order (in this example, the order of times corresponding to the estimation results). More specifically, the display unit 3 calculates each symbol representing the estimation result as a value while arranging the symbols in a predetermined order (in this example, the order of times corresponding to the estimation result) along the horizontal axis. It arrange
- the display unit 3 changes the display position of the cursor 31 according to the movement of the cursor operation unit 7 operated by the observer.
- the display unit 3 displays information representing the learning model corresponding to the estimation result indicated by the symbol in the graph.
- the display state of the graph at this time is illustrated in FIG.
- the display means 3 displays an example in which information representing a learning model corresponding to the estimation result indicated by the symbol is displayed in the graph. I will explain. However, when any symbol enters the on-cursor state, the display unit 3 may display information representing the learning model corresponding to the estimation result indicated by the symbol in the graph.
- FIG. 14 is a flowchart illustrating an example of processing progress of the fourth embodiment.
- Step S41 is the same as step S1 in the first embodiment.
- the display unit 3 arranges the symbols representing the estimation results in a predetermined order along the horizontal axis, and arranges the estimation results calculated as values to be the coordinates in the vertical axis, and the cursor 31. Is displayed (step S42). In step S42, the screen illustrated in FIG. 13A is displayed.
- step S43 the display means 3 determines whether or not any symbol in the graph has been pointed and clicked. If no symbol is point-and-clicked (No in step S43), the determination in step S43 is repeated.
- step S43 the display unit 3 displays information representing the learning model corresponding to the estimation result indicated by the symbol (step S44).
- step S44 the screen illustrated in FIG. 13B is displayed.
- the example shown in FIG. 13B shows an example in which a symbol representing an estimation result derived using “learning model 2” is point-and-clicked and information “learning model 2” is displayed. Yes.
- step S44 when the cursor 31 moves away from the symbol, the display unit 3 may repeat the operations after step S43.
- the display means 3 displays the symbol.
- Information representing a learning model corresponding to the estimation result is displayed. Accordingly, the observer can intuitively recognize at a glance from one graph which learning model was selected when the estimation result represented by the symbol was derived.
- FIG. 15 is a schematic diagram showing a graph in the first modified example of the fourth embodiment.
- the display means 3 displays the cursor 31 together with the graph.
- the display means 3 also displays a legend indicating a plurality of learning models.
- FIG. 15A shows an initial state in which the cursor 31 is displayed together with the graph.
- the display unit 3 changes the display position of the cursor 31 according to the movement of the cursor operation unit 7 operated by the observer. And when a specific learning model is point-and-clicked among the plurality of learning models indicated by the legend displayed in the vicinity of the graph, the display means 3 shows the estimation result derived by the learning model. Highlight the symbol. In other words, the display unit 3 displays a graph in which the display mode (type) of the symbol indicating the estimation result corresponding to the point-and-click learning model and the other symbols are changed. An example of the display state of the graph at this time is shown in FIG.
- the learning model need not be specified by point-and-click.
- the observer may designate a specific learning model by inputting a command through the command line. Alternatively, the observer may select a specific learning model via an interface such as a pull-down menu or a radio button.
- the display means 3 may display a graph (scatter diagram) shown in FIG. 16 as a graph. In this case, the same information as in the second embodiment is input to the input unit 2.
- the display means 3 may determine the symbol arrangement position in the scatter diagram as in the second embodiment. However, the display device 3 shares the types of all symbols. Then, when any symbol is point-and-clicked (or when it becomes an on-cursor state), the display unit 3 displays information representing the learning model corresponding to the estimation result indicated by the symbol. (See FIG. 16). The display unit 3 does not display information representing the learning model when the cursor 31 is away from the symbol.
- the display means 3 may display a graph (scatter diagram) shown in FIG. 17 as a graph. In this case, the same information as that in the third embodiment is input to the input unit 2.
- the display unit 3 may determine the symbol arrangement position in the scatter diagram as in the third embodiment. Further, as in the third embodiment, the display unit 3 changes the type of symbol representing the estimation result according to the estimation result (for example, “city” or “countryside”). However, the display unit 3 does not change the type of the symbol based on the learning model corresponding to the estimation result indicated by the symbol. Then, when any symbol is point-and-clicked (or when it becomes an on-cursor state), the display unit 3 displays information representing the learning model corresponding to the estimation result indicated by the symbol. (See FIG. 17). The display unit 3 does not display information representing the learning model when the cursor 31 is away from the symbol.
- FIG. 5 The estimation result display system according to the fifth embodiment of the present invention can be represented by the block diagram shown in FIG. 5 similarly to the estimation result display system according to the first embodiment. An embodiment will be described. Explanation of matters similar to those in the first embodiment will be omitted as appropriate.
- the estimation result display system 1 includes an input unit 2 and a display unit 3 (see FIG. 5).
- a plurality of sets of information in which an estimation result and information representing a learning model used when deriving the estimation result are associated are input to the input unit 2. .
- the input unit 2 sends the input information to the display unit 3.
- the estimation result is calculated as a value
- Display means 3 displays a graph representing the estimation result as a symbol.
- FIG. 18 is a schematic diagram illustrating an example of a graph displayed by the display unit 3 in the fifth embodiment.
- the horizontal axis of the graph is an axis representing the order of estimation results.
- FIG. 18 illustrates a case where the horizontal axis of the graph represents time. In this case, it suffices if information on the corresponding time is added to the estimation result input to the input unit 2.
- a set of an estimation result and information representing a learning model may be input to the input unit 2 in order of time corresponding to the estimation result.
- the horizontal axis of the graph may not be an axis representing time, but may be an axis representing an input order of a set of an estimation result and information representing a learning model, for example.
- the vertical axis of the graph is the axis corresponding to the estimation result.
- the display means 3 displays a graph in which symbols representing the estimation results are arranged in a predetermined order (in this example, the order of times corresponding to the estimation results). More specifically, the display unit 3 calculates each symbol representing the estimation result as a value while arranging the symbols in a predetermined order (in this example, the order of times corresponding to the estimation result) along the horizontal axis. It arrange
- the display unit 3 uses the symbol representing the estimation result as the x coordinate (coordinate in the horizontal axis) as the order of the estimation result (for example, the order represented by the time), and the estimation result as the y coordinate ( It is arranged at a position to be the coordinate in the vertical axis direction).
- the interval between symbols arranged in order is short, it can be recognized as a line graph.
- FIG. 18 illustrates a case where the interval between symbols arranged in order is short and is recognized as a line graph.
- the operation of the display means 3 described here is the same as the operation of the display means 3 in the first embodiment.
- the display means 3 shares the types of all symbols.
- the display unit 3 displays the area in the graph in a divided manner according to the learning model corresponding to the estimation result indicated by the symbol. Hereinafter, this operation will be described.
- Symbols are arranged in the horizontal axis direction.
- the display unit 3 displays a boundary line perpendicular to the horizontal axis between the two symbols.
- the learning model corresponding to the estimation result indicated by a certain symbol (A) is “learning model 1”.
- the learning model corresponding to the estimation result indicated by the symbol next to symbol A (B) is “learning model 2”.
- the display means 3 displays a boundary line perpendicular to the horizontal axis between the symbols A and B.
- the boundary line is indicated by a broken line.
- the display means 3 may display a boundary line in the vicinity of the symbol that is the end point (see FIG. 18).
- Each symbol in the region sandwiched between two adjacent boundary lines represents an estimation result derived using a common learning model. Accordingly, each region between two adjacent boundary lines corresponds to one learning model.
- the display means 3 displays a region sandwiched between two adjacent boundary lines in a manner corresponding to the learning model corresponding to the region.
- the display unit 3 exemplifies a case in which an identification number representing a learning model corresponding to the region is displayed in a region sandwiched between two adjacent boundary lines.
- the display unit 3 may display a region sandwiched between two adjacent boundary lines with a background color corresponding to the learning model corresponding to the region. For example, the display unit 3 sets the background color of the area corresponding to “learning model 1” to red or the background color of the area corresponding to “learning model 2” to blue, thereby learning corresponding to each area.
- a model may be presented.
- How to change the display mode of each area according to the type of learning model is not particularly limited.
- FIG. 19 is a flowchart illustrating an example of processing progress of the fifth embodiment.
- Step S51 is the same as step S1 in the first embodiment.
- the display unit 3 displays a graph in which the symbols representing the estimation results are arranged in a predetermined order along the horizontal axis, and the estimation results calculated as values are arranged so as to have the coordinates in the vertical axis direction.
- the area in the graph is divided and displayed (step S52).
- the screen illustrated in FIG. 18 is displayed.
- the method for dividing the region in the graph is not particularly limited.
- the display means 3 may classify the regions in the graph by expressing the regions with different colors and patterns.
- the display means 3 displays the area in the graph in a divided manner according to the learning model corresponding to the estimation result indicated by the symbol. Accordingly, the observer can intuitively recognize at a glance from one graph which learning model was selected when the estimation result represented by the symbol was derived.
- FIG. 20 is a schematic diagram showing a graph in a modified example of the fifth embodiment.
- the display means 3 displays a cursor 31 along with the graph.
- An initial state in which the cursor 31 is displayed together with the graph is shown in FIG.
- the estimation result display system 1 also includes a cursor operation means 7.
- the display unit 3 changes the display position of the cursor 31 according to the movement of the cursor operation unit 7 operated by the observer. Then, when a specific learning model is pointed and clicked among a plurality of learning models indicated by the legend displayed in the vicinity of the graph, the display means 3 displays the estimation result derived from the learning model. Highlight the area in the graph that will be displayed. An example of the display state of the graph at this time is shown in FIG. FIG. 20B illustrates a mode in which highlighting is performed by surrounding a region in the graph with a rectangle.
- the learning model need not be specified by point-and-click.
- the observer may designate a specific learning model by inputting a command through the command line. Alternatively, the observer may select a specific learning model via an interface such as a pull-down menu or a radio button.
- FIG. FIG. 21 is a block diagram illustrating a configuration example of an estimation result display system according to the sixth embodiment of this invention.
- the estimation result display system 1A includes an estimation unit 12A, an input unit 2A, and a display unit 3A.
- the estimation unit 12A selects a learning model according to a condition satisfied by the estimation data from a plurality of learning models.
- the estimation means 12A derives an estimation result based on the estimation data and the selected learning model.
- the estimation unit 12A outputs information in which the estimation result is associated with the information representing the learning model used when the estimation result is derived.
- the input means 2A accepts input of information in which an estimation result and information representing a learning model used for deriving the estimation result are associated with each other.
- the input unit 2A may accept input of information further associated with the actual value corresponding to the estimation result and the estimation data used to derive the estimation result.
- the display means 3A displays a graph in which the estimation result is represented by a symbol, and the type of the symbol is changed according to the learning model corresponding to the estimation result.
- the display unit 3A performs, for example, the operations described in the first to sixth embodiments and the modifications thereof.
- the estimation result display system 1 ⁇ / b> A may include a cursor operation unit 7.
- FIG. 22 is a schematic block diagram showing a configuration example of a computer according to each embodiment of the present invention.
- the computer 1000 includes a CPU 1001, a main storage device 1002, an auxiliary storage device 1003, an interface 1004, a display device 1005, and an input device 1006.
- the estimation result display system 1 of each embodiment is implemented in a computer 1000.
- the operation of the estimation result display system 1 is stored in the auxiliary storage device 1003 in the form of a program (estimation result display program).
- the CPU 1001 reads out the program from the auxiliary storage device 1003, develops it in the main storage device 1002, and executes the above processing according to the program.
- the auxiliary storage device 1003 is an example of a tangible medium that is not temporary.
- Other examples of the non-temporary tangible medium include a magnetic disk, a magneto-optical disk, a CD-ROM, a DVD-ROM, and a semiconductor memory connected via the interface 1004.
- this program is distributed to the computer 1000 via a communication line, the computer 1000 that has received the distribution may develop the program in the main storage device 1002 and execute the above processing.
- the program may be for realizing a part of the above-described processing.
- the program may be a differential program that realizes the above-described processing in combination with another program already stored in the auxiliary storage device 1003.
- the first embodiment and the fourth embodiment can be expressed as follows.
- the input unit 2 receives information in which an estimation result is associated with information representing a learning model used in deriving the estimation result.
- the display means 3 is a line graph representing the transition of the estimation result, and displays a graph in which the line attribute is changed according to the learning model corresponding to the estimation result.
- the transition of the estimation result is, for example, a time series transition of the estimation result.
- the attribute of the line is, for example, the appearance of the line such as the color of the line and the type of the line (for example, a solid line, a dotted line, an alternate long and short dash line).
- Line attributes are not limited to line appearance.
- the display unit 3 may display information indicating the learning model corresponding to the line.
- An estimation result display system comprising: a display unit configured to display a graph in which a symbol type is changed according to a learning model corresponding to the estimation result.
- the said display means is an estimated result display system of Additional remark 1 which displays the graph which represented the value of the said estimation result by the coordinate where the said symbol is located.
- the input means includes information in which a value of an estimation result, information representing a learning model used in deriving the estimation result, and an actual value corresponding to the estimation result are associated with each other.
- a set of certain estimation result data is input, and the display means is a scatter diagram having an axis corresponding to the value of the estimation result and an axis corresponding to the actual measurement value, and the estimation result corresponding to the estimation result data
- the estimation result display system according to supplementary note 1 or supplementary note 2, which displays a scatter diagram in which symbols corresponding to the estimation result data are arranged based on the value of and the actual measurement value.
- the said display means is an estimated result display system of Additional remark 1 which displays the graph which represented the said estimation result with the classification of the said symbol.
- the input means includes information indicating an estimation result, values of at least two types of attributes used in deriving the estimation result, and a learning model used in deriving the estimation result
- a set of estimation result data which is information associated with and is input, and the display means has an axis corresponding to a value of a first attribute and an axis corresponding to a value of a second attribute among the attributes.
- the scatter diagram wherein a scatter diagram in which symbols corresponding to the estimation result data are arranged is displayed based on the value of the first attribute and the value of the second attribute corresponding to the estimation result data.
- a set of estimation result data which is information in which an estimation result and information representing a learning model used to derive the estimation result are associated with each other, is input to the input unit, and the display
- the means selects, for each estimation result data, a symbol type representing the estimation result data according to a learning model corresponding to the estimation result data, and selects the estimation result corresponding to the estimation result data as the selected symbol.
- the estimation result display system according to appendix 1, which displays a graph represented by
- An input means for inputting information in which an estimation result is associated with information representing a learning model used to derive the estimation result, and a graph representing the estimation result as a symbol are displayed.
- An estimation result display system comprising: a display unit configured to display a graph and to display a region in the graph according to a learning model corresponding to the estimation result indicated by the symbol.
- An estimation result display system comprising: a display unit configured to display a graph in which line attributes are changed according to a learning model corresponding to the estimation result.
- the estimation result input to the input means is any one of Supplementary notes 1 to 9, which is an estimation result derived from the estimation data and the learning model selected according to the estimation data.
- the learning model according to the conditions which the data for estimation satisfy
- the said estimation The information processing apparatus further includes estimation means for outputting information in which a result is associated with information representing a learning model used when deriving the estimation result, and the input means receives information output from the estimation means as input.
- the estimation result display system according to any one of Supplementary Note 1 to Supplementary Note 10.
- An estimation result display method comprising: displaying a graph in which a symbol type is changed according to a learning model corresponding to an estimation result.
- An estimation result display method characterized by displaying information representing a learning model corresponding to an estimation result indicated by a symbol that has undergone a selection operation when any symbol in the graph is subjected to the selection operation.
- An estimation result display method characterized by displaying a graph in which line attributes are changed according to a learning model corresponding to.
- An estimation result display program mounted on a computer having an input means for inputting information in which an estimation result is associated with information representing a learning model used to derive the estimation result.
- An estimation result for causing the computer to execute a display process for displaying a graph in which the estimation result is represented by a symbol, wherein the graph changes a symbol type according to a learning model corresponding to the estimation result. Display program.
- An estimation result display program mounted on a computer having an input means for inputting information in which an estimation result and information representing a learning model used for deriving the estimation result are input.
- a graph representing the estimation result as a symbol is displayed on the computer, and when any symbol in the graph is subjected to a selection operation, it corresponds to the estimation result indicated by the symbol that has received the selection operation.
- the estimation result display program for performing the display process which displays the information showing the learning model to perform.
- An estimation result display program mounted on a computer having an input means for inputting information in which an estimation result and information representing a learning model used for deriving the estimation result are input. And displaying a graph in which symbols representing the estimation results are arranged in a predetermined order on the computer, and displaying an area in the graph in accordance with a learning model corresponding to the estimation results indicated by the symbols.
- An estimation result display program for executing processing.
- An estimation result display program mounted on a computer having an input means for inputting information in which an estimation result is associated with information representing a learning model used to derive the estimation result.
- An estimation result display program for causing the computer to execute a display process for displaying a graph representing a transition of an estimation result, wherein the graph has a line attribute changed according to a learning model corresponding to the estimation result.
- the present invention is preferably applied to an estimation result display system that selects a learning model and displays an estimation result derived using the learning model.
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Software Systems (AREA)
- General Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Human Computer Interaction (AREA)
- Data Mining & Analysis (AREA)
- Databases & Information Systems (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Evolutionary Computation (AREA)
- Medical Informatics (AREA)
- Artificial Intelligence (AREA)
- Computing Systems (AREA)
- Mathematical Physics (AREA)
- User Interface Of Digital Computer (AREA)
- Information Retrieval, Db Structures And Fs Structures Therefor (AREA)
Abstract
Description
、推定用データがその推定式の説明変数に代入されることによって、推定結果が算出される。ここでは、学習モデルが推定式の形式である場合を例にしたが、学習モデルの形式は推定式であるとは限らない。学習器11によって生成された複数の学習モデルは、推定器12で用いられる。
図5は、本発明の第1の実施形態の推定結果表示システムの構成例を示すブロック図である。本実施形態の推定結果表示システム1は、入力手段2と、表示手段3とを備える。
本発明の第2の実施形態の推定結果表示システムは、第1の実施形態の推定結果表示システムと同様に、図5に示すブロック図で表すことができるので、図5を用いて第2の実施形態を説明する。第1の実施形態と同様の事項については、適宜説明を省略する。
本発明の第3の実施形態の推定結果表示システムは、第1の実施形態の推定結果表示システムと同様に、図5に示すブロック図で表すことができるので、図5を用いて第3の実施形態を説明する。第1の実施形態と同様の事項については、適宜説明を省略する。
図12は、本発明の第4の実施形態の構成例を示すブロック図である。第4の実施形態の推定結果表示システム1は、入力手段2と、表示手段3と、カーソル操作手段7とを備える。なお、第1の実施形態と同様の事項については、適宜説明を省略する。
本発明の第5の実施形態の推定結果表示システムは、第1の実施形態の推定結果表示システムと同様に、図5に示すブロック図で表すことができるので、図5を用いて第5の実施形態を説明する。第1の実施形態と同様の事項については、適宜説明を省略する。
図21は、本発明の第6の実施形態の推定結果表示システムの構成例を示すブロック図である。推定結果表示システム1Aは、推定手段12Aと、入力手段2Aと、表示手段3Aとを含む。
2 入力手段
3 表示手段
7 カーソル操作手段
Claims (19)
- 推定結果と、前記推定結果を導出する際に用いられた学習モデルを表す情報とが対応付けられた情報が入力される入力手段と、
前記推定結果をシンボルで表したグラフであって、前記推定結果に対応する学習モデルに応じてシンボルの種別を変えたグラフを表示する表示手段とを備える
ことを特徴とする推定結果表示システム。 - 前記表示手段は、前記推定結果の値を、前記シンボルが位置する座標によって表したグラフを表示する
請求項1に記載の推定結果表示システム。 - 前記入力手段には、推定結果の値と、前記推定結果を導出する際に用いられた学習モデルを表す情報と、前記推定結果に対応する実測値とが対応付けられた情報である推定結果データの集合が入力され、
前記表示手段は、前記推定結果の値に対応する軸および前記実測値に対応する軸を有する散布図であって、前記推定結果データに対応する前記推定結果の値および前記実測値に基づいて、前記推定結果データに対応するシンボルを配置した散布図を表示する
請求項1または請求項2に記載の推定結果表示システム。 - 前記表示手段は、前記推定結果を、前記シンボルの種別によって表したグラフを表示する
請求項1に記載の推定結果表示システム。 - 前記入力手段には、推定結果と、前記推定結果を導出する際に用いられた少なくとも2種類の属性の値と、前記推定結果を導出する際に用いられた学習モデルを表す情報とが対応付けられた情報である推定結果データの集合が入力され、
前記表示手段は、前記属性のうち第1の属性の値に対応する軸および第2の属性の値に対応する軸を有する散布図であって、前記推定結果データに対応する第1の属性の値および第2の属性の値に基づいて、前記推定結果データに対応するシンボルを配置した散布図を表示する
請求項4に記載の推定結果表示システム。 - 前記入力手段には、推定結果と、前記推定結果を導出する際に用いられた学習モデルを表す情報とが対応付けられた情報である推定結果データの集合が入力され、
前記表示手段は、前記推定結果データ毎に、前記推定結果データに対応する学習モデルに応じて前記推定結果データを表すシンボルの種別を選択し、前記推定結果データに対応する前記推定結果を、選択したシンボルで表したグラフを表示する
請求項1に記載の推定結果表示システム。 - 推定結果と、前記推定結果を導出する際に用いられた学習モデルを表す情報とが対応付けられた情報が入力される入力手段と、
前記推定結果をシンボルで表したグラフを表示するとともに、グラフ内のいずれかのシンボルが選択操作を受けた場合に、当該選択操作を受けたシンボルが示す推定結果に対応する学習モデルを表す情報を表示する表示手段とを備える
ことを特徴とする推定結果表示システム。 - 推定結果と、前記推定結果を導出する際に用いられた学習モデルを表す情報とが対応付けられた情報が入力される入力手段と、
前記推定結果を表すシンボルを所定の順に並べたグラフを表示するとともに、シンボルが示す推定結果に対応する学習モデルに応じて前記グラフ内の領域を区分して表示する表示手段とを備える
ことを特徴とする推定結果表示システム。 - 推定結果と、前記推定結果を導出する際に用いられた学習モデルを表す情報とが対応付けられた情報が入力される入力手段と、
推定結果の推移を表す線グラフであって、推定結果に対応する学習モデルに応じて線の属性を変えたグラフを表示する表示手段とを備える
ことを特徴とする推定結果表示システム。 - 入力手段に入力される推定結果は、推定用データと、前記推定用データに応じて選択された学習モデルとによって導出された推定結果である
請求項1から請求項9のうちのいずれか1項に記載の推定結果表示システム。 - 複数の学習モデルのうちから、推定用データが満たす条件に応じた学習モデルを選択し、前記推定用データと、選択した前記学習モデルとに基づいて推定結果を導出し、前記推定結果と、前記推定結果を導出する際に用いられた学習モデルを表す情報とを対応付けた情報を出力する推定手段をさらに備え、
前記入力手段は、前記推定手段が出力した情報を入力とする
請求項1から請求項10のうちのいずれか1項に記載の推定結果表示システム。 - 推定結果と、前記推定結果を導出する際に用いられた学習モデルを表す情報とが対応付けられた情報の入力を受け付け、
前記推定結果をシンボルで表したグラフであって、前記推定結果に対応する学習モデルに応じてシンボルの種別を変えたグラフを表示する
ことを特徴とする推定結果表示方法。 - 推定結果と、前記推定結果を導出する際に用いられた学習モデルを表す情報とが対応付けられた情報の入力を受け付け、
前記推定結果をシンボルで表したグラフを表示するとともに、グラフ内のいずれかのシンボルが選択操作を受けた場合に、当該選択操作を受けたシンボルが示す推定結果に対応する学習モデルを表す情報を表示する
ことを特徴とする推定結果表示方法。 - 推定結果と、前記推定結果を導出する際に用いられた学習モデルを表す情報とが対応付けられた情報の入力を受け付け、
前記推定結果を表すシンボルを所定の順に並べたグラフを表示するとともに、シンボルが示す推定結果に対応する学習モデルに応じて前記グラフ内の領域を区分して表示する
ことを特徴とする推定結果表示方法。 - 推定結果と、前記推定結果を導出する際に用いられた学習モデルを表す情報とが対応付けられた情報の入力を受け付け、
推定結果の推移を表す線グラフであって、推定結果に対応する学習モデルに応じて線の属性を変えたグラフを表示する
ことを特徴とする推定結果表示方法。 - 推定結果と、前記推定結果を導出する際に用いられた学習モデルを表す情報とが対応付けられた情報が入力される入力手段を備えたコンピュータに搭載される推定結果表示プログラムであって、
前記コンピュータに、
前記推定結果をシンボルで表したグラフであって、前記推定結果に対応する学習モデルに応じてシンボルの種別を変えたグラフを表示する表示処理
を実行させるための推定結果表示プログラム。 - 推定結果と、前記推定結果を導出する際に用いられた学習モデルを表す情報とが対応付けられた情報が入力される入力手段を備えたコンピュータに搭載される推定結果表示プログラムであって、
前記コンピュータに、
前記推定結果をシンボルで表したグラフを表示するとともに、グラフ内のいずれかのシンボルが選択操作を受けた場合に、当該選択操作を受けたシンボルが示す推定結果に対応する学習モデルを表す情報を表示する表示処理
を実行させるための推定結果表示プログラム。 - 推定結果と、前記推定結果を導出する際に用いられた学習モデルを表す情報とが対応付けられた情報が入力される入力手段を備えたコンピュータに搭載される推定結果表示プログラムであって、
前記コンピュータに、
前記推定結果を表すシンボルを所定の順に並べたグラフを表示するとともに、シンボルが示す推定結果に対応する学習モデルに応じて前記グラフ内の領域を区分して表示する表示処理
を実行させるための推定結果表示プログラム。 - 推定結果と、前記推定結果を導出する際に用いられた学習モデルを表す情報とが対応付けられた情報が入力される入力手段を備えたコンピュータに搭載される推定結果表示プログラムであって、
前記コンピュータに、
推定結果の推移を表す線グラフであって、推定結果に対応する学習モデルに応じて線の属性を変えたグラフを表示する表示処理
を実行させるための推定結果表示プログラム。
Priority Applications (3)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
EP15852221.9A EP3211569A4 (en) | 2014-10-21 | 2015-04-30 | Estimation results display system, estimation results display method, and estimation results display program |
JP2016555047A JP6565924B2 (ja) | 2014-10-21 | 2015-04-30 | 推定結果表示システム、推定結果表示方法および推定結果表示プログラム |
US15/521,265 US10885684B2 (en) | 2014-10-21 | 2015-04-30 | Estimation results display system, estimation results display method, and estimation results display program |
Applications Claiming Priority (2)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
JP2014214561 | 2014-10-21 | ||
JP2014-214561 | 2014-10-21 |
Publications (1)
Publication Number | Publication Date |
---|---|
WO2016063433A1 true WO2016063433A1 (ja) | 2016-04-28 |
Family
ID=55760497
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
PCT/JP2015/002297 WO2016063433A1 (ja) | 2014-10-21 | 2015-04-30 | 推定結果表示システム、推定結果表示方法および推定結果表示プログラム |
Country Status (4)
Country | Link |
---|---|
US (1) | US10885684B2 (ja) |
EP (1) | EP3211569A4 (ja) |
JP (1) | JP6565924B2 (ja) |
WO (1) | WO2016063433A1 (ja) |
Cited By (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN109416803A (zh) * | 2016-07-06 | 2019-03-01 | 万事达卡国际公司 | 通过对话界面提供销售信息和见解的方法和系统 |
JPWO2021256442A1 (ja) * | 2020-06-15 | 2021-12-23 | ||
JPWO2021256443A1 (ja) * | 2020-06-15 | 2021-12-23 | ||
RU2808619C1 (ru) * | 2020-06-15 | 2023-11-30 | ДжФЕ СТИЛ КОРПОРЕЙШН | Устройство для измерения механических свойств, способ измерения механических свойств, оборудование для изготовления материала, способ контроля материала и способ изготовления материала |
Families Citing this family (1)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
JP6532762B2 (ja) * | 2015-06-02 | 2019-06-19 | 株式会社東芝 | 情報生成システム、装置、方法、及びプログラム |
Citations (2)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
JP2005135287A (ja) * | 2003-10-31 | 2005-05-26 | National Agriculture & Bio-Oriented Research Organization | 予測装置、予測方法および予測プログラム |
JP2013191099A (ja) * | 2012-03-14 | 2013-09-26 | Fujitsu Ltd | 選択プログラム、選択方法及び選択装置 |
Family Cites Families (6)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
KR100660861B1 (ko) | 2005-02-23 | 2006-12-26 | 삼성전자주식회사 | 반도체 공정 결과를 예측하고 제어하는 반도체 공정 제어장치 |
US8176096B2 (en) * | 2008-12-18 | 2012-05-08 | Microsoft Corporation | Data visualization interactivity architecture |
US8762299B1 (en) * | 2011-06-27 | 2014-06-24 | Google Inc. | Customized predictive analytical model training |
JP5851857B2 (ja) * | 2012-01-20 | 2016-02-03 | アズビル株式会社 | グラフ表示装置 |
US9098942B2 (en) * | 2012-01-25 | 2015-08-04 | Oracle International Corporation | Legend indicator for selecting an active graph series |
US8909582B2 (en) | 2013-02-04 | 2014-12-09 | Nec Corporation | Hierarchical latent variable model estimation device, hierarchical latent variable model estimation method, and recording medium |
-
2015
- 2015-04-30 WO PCT/JP2015/002297 patent/WO2016063433A1/ja active Application Filing
- 2015-04-30 EP EP15852221.9A patent/EP3211569A4/en not_active Withdrawn
- 2015-04-30 JP JP2016555047A patent/JP6565924B2/ja active Active
- 2015-04-30 US US15/521,265 patent/US10885684B2/en active Active
Patent Citations (2)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
JP2005135287A (ja) * | 2003-10-31 | 2005-05-26 | National Agriculture & Bio-Oriented Research Organization | 予測装置、予測方法および予測プログラム |
JP2013191099A (ja) * | 2012-03-14 | 2013-09-26 | Fujitsu Ltd | 選択プログラム、選択方法及び選択装置 |
Non-Patent Citations (2)
Title |
---|
"Ippo Susunda Kikai Gakushu IoT de Gekizo suru Data no Katsuyo Genba ni Shinto", NIKKEI BIG DATA, 10 June 2014 (2014-06-10), pages 7 - 12, XP009503263 * |
See also references of EP3211569A4 * |
Cited By (9)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN109416803A (zh) * | 2016-07-06 | 2019-03-01 | 万事达卡国际公司 | 通过对话界面提供销售信息和见解的方法和系统 |
JPWO2021256442A1 (ja) * | 2020-06-15 | 2021-12-23 | ||
JPWO2021256443A1 (ja) * | 2020-06-15 | 2021-12-23 | ||
WO2021256443A1 (ja) * | 2020-06-15 | 2021-12-23 | Jfeスチール株式会社 | 機械的特性の計測装置、機械的特性の計測方法、物質の製造設備、物質の管理方法および物質の製造方法 |
WO2021256442A1 (ja) * | 2020-06-15 | 2021-12-23 | Jfeスチール株式会社 | 機械的特性の計測装置、機械的特性の計測方法、物質の製造設備、物質の管理方法および物質の製造方法 |
JP7095814B2 (ja) | 2020-06-15 | 2022-07-05 | Jfeスチール株式会社 | 機械的特性の計測装置、機械的特性の計測方法、物質の製造設備、物質の管理方法および物質の製造方法 |
JP7095815B2 (ja) | 2020-06-15 | 2022-07-05 | Jfeスチール株式会社 | 機械的特性の計測装置、機械的特性の計測方法、物質の製造設備、物質の管理方法および物質の製造方法 |
RU2808619C1 (ru) * | 2020-06-15 | 2023-11-30 | ДжФЕ СТИЛ КОРПОРЕЙШН | Устройство для измерения механических свойств, способ измерения механических свойств, оборудование для изготовления материала, способ контроля материала и способ изготовления материала |
RU2808618C1 (ru) * | 2020-06-15 | 2023-11-30 | ДжФЕ СТИЛ КОРПОРЕЙШН | Устройство для измерения механических свойств, способ измерения механических свойств, оборудование для изготовления материала, способ контроля материала и способ изготовления |
Also Published As
Publication number | Publication date |
---|---|
EP3211569A4 (en) | 2018-06-20 |
EP3211569A1 (en) | 2017-08-30 |
US10885684B2 (en) | 2021-01-05 |
JP6565924B2 (ja) | 2019-08-28 |
JPWO2016063433A1 (ja) | 2017-09-07 |
US20170300346A1 (en) | 2017-10-19 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
JP6565925B2 (ja) | 推定結果表示システム、推定結果表示方法および推定結果表示プログラム | |
JP6565924B2 (ja) | 推定結果表示システム、推定結果表示方法および推定結果表示プログラム | |
CN104537355B (zh) | 一种利用图像边界信息和区域连通性的显著对象检测方法 | |
TWI671544B (zh) | 氣象資料處理裝置、包含氣象資料處理裝置之系統、氣象資料處理方法及氣象資料處理程式 | |
JP5658364B2 (ja) | プログラム可視化装置 | |
US20170257291A1 (en) | Node-centric analysis of dynamic networks | |
KR20210048450A (ko) | 차량 주행 경로 지도 생성 방법, 장치 및 컴퓨터프로그램 | |
JP5486939B2 (ja) | 道路交通状況解析システム | |
JP2014524170A5 (ja) | ||
WO2016129218A1 (ja) | 分析用情報表示システム、方法およびプログラム | |
CN102033936A (zh) | 一种用于时间序列相似性比较的方法 | |
US20150338228A1 (en) | Route planning method and route planning system | |
WO2016132684A1 (ja) | 影響度可視化システム、方法およびプログラム | |
JP6029843B2 (ja) | 地図表示装置 | |
US20150362548A1 (en) | Wafer map identification system for wafer test data | |
JP5197790B2 (ja) | 運転曲線作成装置および運転曲線作成方法 | |
WO2015159377A1 (ja) | 設計支援装置 | |
US20170068893A1 (en) | System, Method and Software for Representing Decision Trees | |
WO2019092778A1 (ja) | 表示プログラム、表示方法及び表示装置 | |
JP6165001B2 (ja) | 波形描画装置及び波形描画プログラム | |
CN110348495A (zh) | 一种改进K-means聚类的航迹关联方法 | |
KR102230814B1 (ko) | 차량 주행 경로 지도 생성 방법, 장치 및 컴퓨터프로그램 | |
CN104298440B (zh) | 信息处理装置 | |
JP2018049565A (ja) | ソフトウエア検証用システム及びソフトウエア検証方法 | |
CN106371400A (zh) | 一种示教轨迹选中方法及装置 |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
121 | Ep: the epo has been informed by wipo that ep was designated in this application |
Ref document number: 15852221 Country of ref document: EP Kind code of ref document: A1 |
|
REEP | Request for entry into the european phase |
Ref document number: 2015852221 Country of ref document: EP |
|
ENP | Entry into the national phase |
Ref document number: 2016555047 Country of ref document: JP Kind code of ref document: A |
|
NENP | Non-entry into the national phase |
Ref country code: DE |
|
WWE | Wipo information: entry into national phase |
Ref document number: 15521265 Country of ref document: US |