U.S. patent application number 11/519317 was filed with the patent office on 2007-01-11 for system and method for mining model accuracy display.
This patent application is currently assigned to Microsoft Corporation. Invention is credited to David Earl Heckerman, Pyungchul Kim, Scott Conrad Oveson, Zhaohui Tang.
Application Number | 20070010966 11/519317 |
Document ID | / |
Family ID | 29779802 |
Filed Date | 2007-01-11 |
United States Patent
Application |
20070010966 |
Kind Code |
A1 |
Kim; Pyungchul ; et
al. |
January 11, 2007 |
System and method for mining model accuracy display
Abstract
Systems and methods are provided for producing displays of the
accuracy of data mining or statistical models that produce
associative predictions. For all cases in a testing data set, the
model makes predictions and provides associated probabilities. The
cases are sorted by their probability of making accurate
predictions and a graph is made of the accuracy of the model over
various subsets containing the highest probability cases as
evaluated by the model. Where a number of probabilities are
presented for the predictions in a basket of predictions, those
probabilities are combined to yield a probability score for the
entire basket. Additionally, the accuracy of a model over different
basket sizes may be graphed. The accuracy graph may also be
produced for any models making a prediction, by graphing the
probability of making accurate predictions and a graph made of the
accuracy of the model over various subsets of the data containing
the highest probability cases.
Inventors: |
Kim; Pyungchul; (Snoqualmie,
WA) ; Tang; Zhaohui; (Bellevue, WA) ;
Heckerman; David Earl; (Bellevue, WA) ; Oveson; Scott
Conrad; (Sammamish, WA) |
Correspondence
Address: |
WOODCOCK WASHBURN LLP (MICROSOFT CORPORATION)
ONE LIBERTY PLACE - 46TH FLOOR
PHILADELPHIA
PA
19103
US
|
Assignee: |
Microsoft Corporation
Redmond
WA
98052
|
Family ID: |
29779802 |
Appl. No.: |
11/519317 |
Filed: |
September 11, 2006 |
Related U.S. Patent Documents
|
|
|
|
|
|
Application
Number |
Filing Date |
Patent Number |
|
|
10186052 |
Jun 28, 2002 |
7124054 |
|
|
11519317 |
Sep 11, 2006 |
|
|
|
Current U.S.
Class: |
702/181 |
Current CPC
Class: |
G06F 16/2465 20190101;
G06F 17/18 20130101 |
Class at
Publication: |
702/181 |
International
Class: |
G06F 17/18 20060101
G06F017/18; G06F 19/00 20060101 G06F019/00 |
Claims
1. A method for accuracy display for a model, where, for each case
in a testing data set containing one or more associated data items,
said model predicts a basket containing a number of predicted data
items associated with said case comprising: producing a visual
representation of a comprehensive association prediction evaluation
on a graph of at least two dimensions, said comprehensive
association prediction evaluation comprising at least two points
where for each of said data points: a second coordinate of said
data point corresponds to the accuracy of said model when
predicting a basket containing a number of items corresponding to a
first coordinate of said data point.
2. A method according to claim 1, wherein said accuracy of said
model when predicting a basket containing a number of items
corresponding to a first coordinate of said data point corresponds
to the number of cases for which said determination determines that
said one of said associated data items removed from said case is
contained in said basket predicted by said model.
3. A method according to claim 1, wherein said accuracy of said
model when predicting a basket containing a number of items
corresponding to a first coordinate of said data point corresponds
to the proportion of cases from among said testing data set for
which said determination determines that said one of said
associated data items removed from said case is contained in said
basket predicted by said model.
4. A method according to claim 1, where at a visual representation
of a comprehensive association prediction evaluation for at least
one other model is displayed on said graph.
5. A modulated data signal carrying computer executable
instructions for performing the method of claim 1.
6. A computer-readable medium comprising computer-executable
modules having computer-executable instructions for accuracy
display for a model, where, for each case in a testing data set
containing one or more associated data items, said model predicts a
basket containing a number of predicted data items associated with
said case, said modules comprising: a module for producing a visual
representation of a comprehensive association prediction evaluation
on a graph of at least two dimensions, said comprehensive
association prediction evaluation comprising at least two points
where for each of said data points: a second coordinate of said
data point corresponds to the accuracy of said model when
predicting a basket containing a number of items corresponding to a
first coordinate of said data point.
7. A computer-readable medium according to claim 6, wherein said
accuracy of said model when predicting a basket containing a number
of items corresponding to a first coordinate of said data point
corresponds to the number of cases for which said determination
determines that said one of said associated data items removed from
said case is contained in said basket predicted by said model.
8. A computer-readable medium according to claim 6, wherein said
accuracy of said model when predicting a basket containing a number
of items corresponding to a first coordinate of said data point
corresponds to the proportion of cases from among said testing data
set for which said determination determines that said one of said
associated data items removed from said case is contained in said
basket predicted by said model.
9. A computer-readable medium according to claim 6, where at a
visual representation of a comprehensive association prediction
evaluation for at least one other model is displayed on said
graph.
10. A computer device for accuracy display for a model, where, for
each case in a testing data set containing one or more associated
data items, said model predicts a basket containing a number of
predicted data items associated with said case, comprising: means
producing a visual representation of a comprehensive association
prediction evaluation on a graph of at least two dimensions, said
comprehensive association prediction evaluation comprising at least
two points where for each of said data points: a second coordinate
of said data point corresponds to the accuracy of said model when
predicting a basket containing a number of items corresponding to a
first coordinate of said data point.
11. A computer device according to claim 10, wherein said accuracy
of said model when predicting a basket containing a number of items
corresponding to a first coordinate of said data point corresponds
to the number of cases for which said determination determines that
said one of said associated data items removed from said case is
contained in said basket predicted by said model.
12. A computer device according to claim 10, wherein said accuracy
of said model when predicting a basket containing a number of items
corresponding to a first coordinate of said data point corresponds
to the proportion of cases from among said testing data set for
which said determination determines that said one of said
associated data items removed from said case is contained in said
basket predicted by said model.
13. A computer device according to claim 10, where at a visual
representation of a comprehensive association prediction evaluation
for at least one other model is displayed on said graph.
Description
CROSS-REFERENCE TO RELATED APPLICATIONS
[0001] This application is a continuation of U.S. patent
application Ser. No. 10/186,052, filed on Jun. 28, 2002.
FIELD OF THE INVENTION
[0002] The present invention relates to systems and methods for
evaluating and displaying the reliability of data mining or
statistical models. The present invention also relates to systems
and methods for displaying the accuracy of models that predict
attribute association and for calculating a prediction score for
multiple associative predictions. More particularly, the present
invention relates to systems and methods for displaying a graph
that describes the accuracy of predictive models, and systems and
methods for displaying a graph that describes the accuracy of one
or more attribute association models and to calculating a
prediction score for combining the composite probability of several
associative predictions, each with their own probability value.
BACKGROUND OF THE INVENTION
[0003] Data mining is the exploration and analysis of large
quantities of data, in order to discover correlations, patterns,
and trends in the data. Data mining may also be used to create
models that can be used to predict future data or classify existing
data.
[0004] For example, a business may amass a large collection of
information about its customers. This information may include
purchasing information and any other information available to the
business about the customer. The predictions of a model associated
with customer data may be used, for example, to control customer
attrition, to perform credit-risk management, to detect fraud, or
to make decisions on marketing.
[0005] To create and test a data mining model, available data may
be divided into two parts. One part, the training data set, may be
used to create models. The rest of the data, the testing data set,
may be used to test the model, and thereby determine the accuracy
of the model in making predictions.
[0006] Data within data sets is grouped into cases. For example,
with customer data, each case corresponds to a different customer.
Data in the case describes or is otherwise associated with that
customer. One type of data that may be associated with a case (for
example, with a given customer) is a categorical variable. A
categorical variable categorizes the case into one of several
pre-defined states. For example, one such variable may correspond
to the educational level of the customer. There are various values
for this variable. The possible values are known as states. The
states of the educational level variable may be "high school
degree," "bachelor's degree," or "graduate degree" and may
correspond to the highest degree earned by the customer.
[0007] Data available is partitioned into two groups--a training
data set and a testing data set. Often 70% of the data is used for
training and 30% for testing. A model may be trained using only the
training data set, which includes the state information. Once a
model is trained, it may be run on the testing data set for
evaluation. During this testing, the model will be given all of the
testing data except the educational level data, and asked to
predict a probability that the educational level variable for that
customer is a particular state, such as "bachelor's degree".
[0008] Running the model on the testing data set, these results are
compared to the actual testing data to see whether the model
correctly predicted a high probability of the "bachelor's degree"
state for cases that actually have "bachelor's degree" as the state
of the educational level variable.
[0009] One method of displaying the success of a model graphically
is by means of a lift chart, also known as a cumulative gains
chart. To create a lift chart, the cases from the testing data set
are sorted according to the probability assigned by the model that
the variable (e.g. educational level) has the state (e.g.
bachelor's degree) that was tested, from highest probability to
lowest probability. Once this is done, a lift chart can be created
from data points (X, Y) showing for each point what number Y of the
total number of true positives (those cases where the variable does
have the state being tested for) are included in the X % of the
testing data set cases with the highest probability for that state,
as assigned by the model.
[0010] As shown in FIG. 1, the conventional lift chart shows that
there are 1000 total true positives in the testing set. This is not
necessarily the number of cases in the testing data set. Some cases
may have a different state for the variable than the one being
tested. The number of true positives in the testing data set is the
highest number shown on Y axis 10. The X axis 20 correlates with
the percentage of cases with the highest probabilities. Lift line
30 depicts the success of the model. For example, it can be seen
that lift line 30 includes a point with (X, Y) coordinates are
approximately (20, 500). This indicates that, in the 20% of the
cases selected by the model as the most probable cases having the
tested-for state of the variable, approximately 500 of the cases
that are truly positive for the state of the variable are included.
This is equivalent to getting 50% of the actual cases with the
desired state in only 20% of the cases tested for.
[0011] A model that randomly assigns probabilities would be likely
to have a chart close to the random lift line 40. In the top 10% of
cases, such a model would find 10% of the true positives. Note that
the X axis may also be expressed in the number of high probability
cases, and the Y axis in percentages. A perfect model may also be
considered. In a situation where there are N % true positives among
the entire testing data set the lift line would stretch straight
from the origin to the point (N, Y.sub.MAX) (where Y.sub.MAX is the
maximum Y value). This is because all of the true positives would
be identified before any false positives are identified. The lift
line for the perfect model would then continue horizontally from
that point to the right. For example, if 20% of the cases had the
tested for state, as shown in FIG. 2, a perfect model would have
the perfect lift line 50, extending from (0,0) to (20, 1000) and
then from (20, 1000) to (100, 1000). Similarly, the worst case
model would identify no true positives until the last N % of the
testing population is included, and, as shown in FIG. 3 for the
case where there are 20% true positives, the worst case lift line
60 for such a model would extend from (0,0) to (80, 0) and then
straight from (80,0) to (100, 1000).
[0012] As described above, in the prior art, a lift chart can be
used to display and measure the prediction accuracy of a model for
a given state of a categorical variable. However, existing lift
charts do not have any capability for measuring the effectiveness
of a model in predicting an association. Additionally, the prior
art lift chart can be used to display the prediction accuracy of a
model in terms of the percentage of true positives captured in
different size groups of cases with the highest associated
probabilities, however, there is no capability for understanding
what the size of the number of true positives in the testing data
set.
[0013] Thus, there is a need for a method and system for generating
for display improved charts with which to display the accuracy of
models.
SUMMARY OF THE INVENTION
[0014] In view of the foregoing, the present invention provides
systems and methods for creating a chart displaying the accuracy of
a model in making predictions. The present invention also provides
systems and methods for creating a chart displaying the accuracy of
a model in predicting associations. Graphical displays of the
accuracy of the model in single-association and
multiple-association predictions are provided, as well as a
technique for graphically displaying the predictive value of a
model across varying numbers of suggested items.
[0015] According to one embodiment of the inventive technique, one
piece of association data is removed from each case in the testing
data set. The model is used to make a prediction of an association
that matches the remaining data in each case. The model also
provides a probability for the prediction. This prediction and
probability, along with knowledge of whether the predicted
association was, indeed, the association removed from the case, can
be used to create a chart. The cases are sorted by probability, and
the number of correct predictions Y in the most probable X percent
of the cases is graphed.
[0016] However, if there are a large number of possible
associations, most models given the chance to make one association
will not perform very well. For example, where there are thousands
of products that may have been purchased by the consumer, a model
may not be successful in determining the correct product given only
one chance. Therefore, results from a chart tracking a model's
success in making one association given one chance may, in some
cases, not provide significant useful information. In such cases,
allowing a model to select a basket containing multiple predictions
in order to identify the missing association may provide more
useful information about the model. The present invention,
therefore, also presents a technique for displaying the accuracy of
a model providing multiple predictions in graphical form. Again,
one association is removed from each case in the testing data set.
A basket of multiple predictions are made by the model, and the
model assigns a cumulative probability to the basket of
predictions. Where individual probabilities from each prediction in
the basket are assigned, a probability score is created from these
probabilities. The cases are sorted by the probability score or
cumulative probability, and the number of correct predictions Y in
the most probable X percent of the cases is graphed.
[0017] Additionally, a technique for displaying the accuracy of a
model that predicts associations in a novel association prediction
accuracy chart is provided. A comprehensive association accuracy
line is created from points such that, for each point (X,Y), X
corresponds to the size of a prediction basket. For example, if the
associations are products that a customer has purchased or might
purchase and there are N products, the X axis would range between
zero and N. Other ranges may also be used. The Y value for each
data point corresponds to the model's effectiveness when making X
predictions. The Y value is calculated by allowing the model to
make a basket of X predictions for each case, and calculating the
percentage of total cases in which the model picked the missing
association in the basket. When various (X,Y) pairs are calculated
and graphed, a multiple-association accuracy line that displays the
effectiveness of the model is produced.
[0018] Additionally, a technique for displaying the accuracy of a
model that makes predictions regarding data cases is provided. A
comprehensive association accuracy line is created from points such
that, for each point (X,Y), the number of correct predictions Y in
the most probable X percent of the cases is graphed.
[0019] Other features and embodiments of the present invention are
described below.
BRIEF DESCRIPTION OF THE DRAWINGS
[0020] The system and methods for model accuracy display for
multiple state prediction in accordance with the present invention
are further described with reference to the accompanying drawings
in which:
[0021] FIG. 1 is a lift chart according to the prior art with an
exemplary model lift line and a random lift line depicted.
[0022] FIG. 2 is a lift chart according to the prior art with an
exemplary model lift line and an ideal lift line depicted.
[0023] FIG. 3 is a lift chart according to the prior art with an
exemplary model lift line and a worst case lift line depicted.
[0024] FIG. 4 is a block diagram of an exemplary computing
environment in which aspects of the invention may be
implemented.
[0025] FIG. 5 is a chart according to the present invention with a
single-association prediction evaluation line depicted.
[0026] FIG. 6 is a chart according to the present invention with a
multiple-association prediction evaluation line depicted.
[0027] FIG. 7 is a chart according to the present invention with
several association prediction evaluation lines depicted.
[0028] FIG. 8 is a chart according to the present invention with a
comprehensive association prediction accuracy line depicted.
[0029] FIG. 9 is a flow chart of a method for displaying accuracy
of a model in making predictions according to the present
invention.
DETAILED DESCRIPTION OF THE INVENTION
Overview
[0030] As described in the background, the conventional lift chart
can only display the effectiveness of a model at predicting one
state of a multi-state variable. A method and system are presented
for providing a display of the accuracy of a data-mining model that
predicts associations given information in the testing data set,
rather than predicting states of a categorical variable. The method
and system will be described with reference to a data mining model,
however, any statistical model that can be used to make associative
predictions may be used with the method and system of the
invention.
[0031] One associated piece of data is removed from each case in
the testing data set. The model is requested to make one
prediction, or, in an alternative embodiment, a basket of several
predictions, as to what data is associated with the remaining data
in the case. The model also provides a probability for each
prediction. In the case where the model provides multiple
predictions with separate probabilities, a combined probability
score may be calculated from individual probabilities. The cases
are sorted by the associated probability or by the combined
probability score.
[0032] The cases are then examined to determine whether the
prediction or predictions contain the correct association. The
percentage of correct predictions among the highest-probability
cases is graphed.
[0033] Additionally, a comprehensive association prediction
accuracy chart may be prepared. This chart graphs points (X,Y)
corresponding to the effectiveness of the model Y across all cases
when the model has made a number X of associative predictions.
[0034] Additionally, an accuracy chart may be created for any model
making predictions on a number of cases in a data set. Once the
model has made predictions and assigned associated probabilities to
the predictions, the cases are then examined to determine whether
the predictions are correct. The amount of correct predictions
among the highest-probability cases is graphed.
Exemplary Computing Environment
[0035] FIG. 4 illustrates an example of a suitable computing system
environment 100 in which the invention may be implemented. The
computing system environment 100 is only one example of a suitable
computing environment and is not intended to suggest any limitation
as to the scope of use or functionality of the invention. Neither
should the computing environment 100 be interpreted as having any
dependency or requirement relating to any one or combination of
components illustrated in the exemplary operating environment
100.
[0036] One of ordinary skill in the art can appreciate that a
computer or other client or server device can be deployed as part
of a computer network, or in a distributed computing environment.
In this regard, the present invention pertains to any computer
system having any number of memory or storage units, and any number
of applications and processes occurring across any number of
storage units or volumes, which may be used in connection with the
present invention. The present invention may apply to an
environment with server computers and client computers deployed in
a network environment or distributed computing environment, having
remote or local storage. The present invention may also be applied
to standalone computing devices, having programming language
fimctionality, interpretation and execution capabilities for
generating, receiving and transmitting information in connection
with remote or local services.
[0037] The invention is operational with numerous other general
purpose or special purpose computing system environments or
configurations. Examples of well known computing systems,
environments, and/or configurations that may be suitable for use
with the invention include, but are not limited to, personal
computers, server computers, hand-held or laptop devices,
multiprocessor systems, microprocessor-based systems, set top
boxes, programmable consumer electronics, network PCs,
minicomputers, mainframe computers, distributed computing
environments that include any of the above systems or. devices, and
the like.
[0038] The invention may be described in the general context of
computer-executable instructions, such as program modules, being
executed by a computer. Generally, program modules include
routines, programs, objects, components, data structures, etc. that
perform particular tasks or implement particular abstract data
types. The invention may also be practiced in distributed computing
environments where tasks are performed by remote processing devices
that are linked through a communications network or other data
transmission medium. In a distributed computing environment,
program modules and other data may be located in both local and
remote computer storage media including memory storage devices.
Distributed computing facilitates sharing of computer resources and
services by direct exchange between computing devices and systems.
These resources and services include the exchange of information,
cache storage, and disk storage for files. Distributed computing
takes advantage of network connectivity, allowing clients to
leverage their collective power to benefit the entire enterprise.
In this regard, a variety of devices may have applications, objects
or resources that may utilize the techniques of the present
invention.
[0039] With reference to FIG. 4, an exemplary system for
implementing the invention includes a general-purpose computing
device in the form of a computer 110. Components of computer 110
may include, but are not limited to, a processing unit 120, a
system memory 130, and a system bus 121 that couples various system
components including the system memory to the processing unit 120.
The system bus 121 may be any of several types of bus structures
including a memory bus or memory controller, a peripheral bus, and
a local bus using any of a variety of bus architectures. By way of
example, and not limitation, such architectures include Industry
Standard Architecture (ISA) bus, Micro Channel Architecture (MCA)
bus, Enhanced ISA (EISA) bus, Video Electronics Standards
Association (VESA) local bus, and Peripheral Component Interconnect
(PCI) bus (also known as Mezzanine bus).
[0040] Computer 110 typically includes a variety of computer
readable media. Computer readable media can be any available media
that can be accessed by computer 110 and includes both volatile and
nonvolatile media, removable and non-removable media. By way of
example, and not limitation, computer readable media may comprise
computer storage media and communication media. Computer storage
media includes both volatile and nonvolatile, removable and
non-removable media implemented in any method or technology for
storage of information such as computer readable instructions, data
structures, program modules or other data. Computer storage media
includes, but is not limited to, RAM, ROM, EEPROM, flash memory or
other memory technology, CDROM, digital versatile disks (DVD) or
other optical disk storage, magnetic cassettes, magnetic tape,
magnetic disk storage or other magnetic storage devices, or any
other medium that can be used to store the desired information and
that can accessed by computer 110. Communication media typically
embodies computer readable instructions, data structures, program
modules or other data in a modulated data signal such as a carrier
wave or other transport mechanism and includes any information
delivery media. The term "modulated data signal" means a signal
that has one or more of its characteristics set or changed in such
a manner as to encode information in the signal. By way of example,
and not limitation, communication media includes wired media such
as a wired network or direct-wired connection, and wireless media
such as acoustic, RF, infrared and other wireless media.
Combinations of any of the above should also be included within the
scope of computer readable media.
[0041] The system memory 130 includes computer storage media in the
form of volatile and/or nonvolatile memory such as read only memory
(ROM) 131 and random access memory (RAM) 132. A basic input/output
system 133 (BIOS), containing the basic routines that help to
transfer information between elements within computer 110, such as
during start-up, is typically stored in ROM 131. RAM 132 typically
contains data and/or program modules that are immediately
accessible to and/or presently being operated on by processing unit
120. By way of example, and not limitation, FIG. 4 illustrates
operating system 134, application programs 135, other program
modules 136, and program data 137.
[0042] The computer 110 may also include other
removable/non-removable, volatile/nonvolatile computer storage
media. By way of example only, FIG. 4 illustrates a hard disk drive
140 that reads from or writes to non-removable, nonvolatile
magnetic media, a magnetic disk drive 151 that reads from or writes
to a removable, nonvolatile magnetic disk 152, and an optical disk
drive 155 that reads from or writes to a removable, nonvolatile
optical disk 156, such as a CD ROM or other optical media. Other
removable/non-removable, volatile/nonvolatile computer storage
media that can be used in the exemplary operating environment
include, but are not limited to, magnetic tape cassettes, flash
memory cards, digital versatile disks, digital video tape, solid
state RAM, solid state ROM, and the like. The hard disk drive 141
is typically connected to the system bus 121 through an
non-removable memory interface such as interface 140, and magnetic
disk drive 151 and optical disk drive 155 are typically connected
to the system bus 121 by a removable memory interface, such as
interface 150.
[0043] The drives and their associated computer storage media
discussed above and illustrated in FIG. 4, provide storage of
computer readable instructions, data structures, program modules
and other data for the computer 110. In FIG. 4, for example, hard
disk drive 141 is illustrated as storing operating system 144,
application programs 145, other program modules 146, and program
data 147. Note that these components can either be the same as or
different from operating system 134, application programs 135,
other program modules 136, and program data 137. Operating system
144, application programs 145, other program modules 146, and
program data 147 are given different numbers here to illustrate
that, at a minimum, they are different copies. A user may enter
commands and information into the computer 20 through input devices
such as a keyboard 162 and pointing device 161, commonly referred
to as a mouse, trackball or touch pad. Other input devices (not
shown) may include a microphone, joystick, game pad, satellite
dish, scanner, or the like. These and other input devices are often
connected to the processing unit 120 through a user input interface
160 that is coupled to the system bus, but may be connected by
other interface and bus structures, such as a parallel port, game
port or a universal serial bus (USB). A monitor 191 or other type
of display device is also connected to the system bus 121 via an
interface, such as a video interface 190. In addition to the
monitor, computers may also include other peripheral output devices
such as speakers 197 and printer 196, which may be connected
through an output peripheral interface 190.
[0044] The computer 110 may operate in a networked environment
using logical connections to one or more remote computers, such as
a remote computer 180. The remote computer 180 may be a personal
computer, a server, a router, a network PC, a peer device or other
common network node, and typically includes many or all of the
elements described above relative to the computer 110, although
only a memory storage device 181 has been illustrated in FIG. 4.
The logical connections depicted in FIG. 4 include a local area
network (LAN) 171 and a wide area network (WAN) 173, but may also
include other networks. Such networking environments are
commonplace in offices, enterprise-wide computer networks,
intranets and the Internet.
[0045] When used in a LAN networking environment, the computer 110
is connected to the LAN 171 through a network interface or adapter
170. When used in a WAN networking environment, the computer 110
typically includes a modem 172 or other means for establishing
communications over the WAN 173, such as the Internet. The modem
172, which may be internal or external, may be connected to the
system bus 121 via the user input interface 160, or other
appropriate mechanism. In a networked environment, program modules
depicted relative to the computer 110, or portions thereof, may be
stored in the remote memory storage device. By way of example, and
not limitation, FIG. 4 illustrates remote application programs 185
as residing on memory device 181. It will be appreciated that the
network connections shown are exemplary and other means of
establishing a communications link between the computers may be
used.
Calculation and Display of Associative Prediction Evaluation
[0046] In order to calculate and display an evaluation of the
success of a model in predicting an association, the inventive
technique evaluates whether the model can correctly identify an
item that has been removed from a test case.
[0047] In order to evaluate the accuracy of an associative model,
the testing data set used must contain cases that include
associated data. For example, the model can be used on a testing
data set where each case describes one customer, and where each
case contains data about purchases by that customer.
[0048] The technique displays an evaluation of the success of the
model by first removing one item of associated data from each case
in the testing data set. Then the model can either be tested on a
single-association prediction or multiple-association (basket)
prediction, with the results displayed in graphical form.
[0049] The model's accuracy may also be evaluated by generating for
display a comprehensive association prediction accuracy chart which
displays the accuracy of the model across a range of numbers of
predictions made.
Single-Association Testing and Display
[0050] Single-association testing is the testing of the model to
see if it can, with only one prediction, correctly predict which
item of associated data has been removed from each case in the
testing data set.
[0051] Once a case has had an item removed, the model is queried
for a prediction regarding an item associated with the remaining
items in the case. The model also provides the probability that it
assigns to the correctness of its prediction.
[0052] The cases may then be ordered with respect to the
probabilities assigned to the prediction for each case. The
predictions are checked to determine whether they were correct.
This data may then be displayed on a chart. As shown in FIG. 5, a
single-association prediction line 530 is generated from points
(X,Y), where for a group of cases with the highest probabilities
assigned by the model, X corresponds to the number of cases in that
group and Y corresponds to the number of correct predictions in
that group. In the exemplary graph in FIG. 5, the X axis 520 is
measured in percentages, and the Y axis 510 in total number of
cases. The maximum value on the Y axis 510 is 1000. This may be the
total number of cases in the testing data set or some smaller
number of cases. A Y axis with a maximum less than 100% or the
number of total cases may be used for a clearer display when the
total number of correct identifications is small.
Multiple-Association Testing and Lift Chart Display
[0053] A model may also have the capability to provide more than
one prediction for the missing associated data item. This multiple
prediction is also called a basket prediction. In this case, if the
model also provides a composite probability for its set of
predictions, the composite probability can be used according to the
method described for a single association prediction.
[0054] If, however, the model only provides a probability for each
of its predictions for the missing associated data item, a combined
probability score must be assigned. Where the model gives a
probability score for each possible item, in one embodiment the
formula for a combined probability score is given by: i = 1 k
.times. p .function. ( A i ) j = k + 1 N .times. p .function. ( A j
) ( 1 ) ##EQU1## In this formula (1): [0055] k represents the
number of predictions made by the model; [0056] N represents the
total number of association items which may be predicted by a
model; [0057] A.sub.i represents association item i, where items 1
through k are the predictions made by the model, and k+1 through N
are the association items not included in the predictions made by
the model; and [0058] p(A.sub.i) represents the probability that
the model assigns to A.sub.i being the held out item.
[0059] Other formulas for a probability score, including prior art
formulas for combining probabilities, are possible.
[0060] Once a probability score is assigned to the basket of
predictions made by the model, the cases may be ordered with
respect to these probability scores. If the model assigns a single
probability to a basket of predictions, this probability is used to
order the cases. The predictions are checked to determine whether
the removed item was identified in the basket of predictions. If
so, the basket of predictions is deemed to be correct. This data
may be graphed.
[0061] As shown in FIG. 6, a multiple-association prediction line
630 is generated from points (X,Y), where for a group of cases with
the highest probability score (or probability assigned by the
model), X is the amount of cases in that group and Y is the amount
of correct predictions in that group. In the exemplary graph in
FIG. 6, the X axis 620 is measured in percentages, and the Y axis
610 in total number. As before, the maximum value on the Y axis is
1000, which may represent the total number of cases in the testing
data set or some smaller number of cases. Again, a Y axis with a
maximum less than 100% or the number of total cases may be used for
a clearer display when the total number of correct identifications
is small.
Calculation and Display of Associative Prediction Accuracy
Graph
[0062] The single-association prediction line and the
multiple-association prediction line displays described above
provide an accuracy display for the model when it makes a set
number k of predictions per case. It can be appreciated that the
single-association prediction graph is equivalent to a
multiple-association prediction graph where the size of the basket
is one. Thus, for the single-association prediction line, k=1. In
order to determine the model's effectiveness for a number of values
of k, different graphs must be superimposed on each other. This is
shown in FIG. 7, where the single-association prediction line 530
of FIG. 5 and the multiple-association prediction line 630 of FIG.
7are displayed, along with a second multiple-association prediction
line 730. X axis 720 and Y axis 710 provide a common reference for
the prediction lines 530, 630, and 730. In this way, it can be seen
how a number of such association prediction lines can be used to
describe and compare model accuracy over several values of k.
Additionally, prediction accuracy lines for the same basket size
evaluating the accuracy of different models may also be displayed
on one graph in order to compare the accuracy of several
models.
[0063] In addition to displays with multiple lines for multiple
basket sizes, the accuracy of a model over several values of k may
also be graphed. To produce this type of graph, a comprehensive
association prediction accuracy graph, the accuracy of a model over
all cases given a certain basket size is noted. This becomes a
point from which a line may be generated. As shown in FIG. 8, the X
axis 820 defines basket sizes, and the Y axis 810 defines percent
of correct predictions. A point (X, Y) is graphed where, for all
cases, the model predicts Y % of the cases correctly for a basket
size of X. The result is comprehensive association accuracy
prediction line 830. The X axis 820 may range over all possible
basket sizes or may define a subrange of basket sizes to analyze.
In the exemplary graph of FIG. 8, all possible basket sizes are
included. It is notable that where X equals 0 (no predictions are
made) Y also equals 0 (no correct predictions are made for any
cases). Where X equals a basket size containing all possible
predictions, Y equals 100% (or all cases identified correctly).
This is because in a basket of all possible predictions, the
correct prediction will always be included. A comprehensive
association accuracy prediction line may be produced for several
different models, and the models thereby compared.
Calculation and Display of Model Accuracy
[0064] The new accuracy display technique is also applicable more
generally. As shown in FIG. 9, given a model which can make a
prediction, in order to produce an accuracy display, the model
first makes a prediction and produces an associated probability for
the prediction for each case in a testing data set 900. The
correctness of the predictions is determined 910. A graph is then
produced of the percentage of correct predictions for different
groups of the cases with the highest probabilities. For example,
the graph can be made from points (X,Y), where X corresponds to a
percentage of cases from the testing data set, and where Y
corresponds to the accuracy of predictions among the X % of the
cases with the highest probabilities 920. Again, the X and Y axes
may be measured in percentages, numbers of cases, or any similar
measure of amount.
[0065] All of the prediction evaluation lines of the invention may
be produced using some approximations. Not all points (X,Y) on the
line must be exact, and the line may be produced using prior art
algorithms for creating a representative line from data points. In
this way, computational time may be saved for a small cost in
accuracy of the display line. In place of lines, data points may be
displayed. Equivalent graphs may be produced as is known in the
prior art, by changing the scale of the axes, or by changing the
position of the axes.
[0066] These and other possible variations that would be obvious to
one skilled in the art are contemplated, and the invention should
not be limited to any single embodiment.
CONCLUSION
[0067] Herein a system and method for mining model accuracy display
for associative prediction is provided that produces a display of
the accuracy of a model in providing associative predictions.
[0068] The invention also contemplates placing more than
associative prediction evaluation lines on a single graph in order
to compare and contrast the accuracy of one model over different
basket sizes, to compare and contrast the accuracy of more than one
model, or to compare and contrast the accuracy of one model on
several testing data sets, which may have been selected for various
attributes of the data sets.
[0069] As mentioned above, while exemplary embodiments of the
present invention have been described in connection with various
computing devices and network architectures, the underlying
concepts may be applied to any computing device or system in which
it is desirable to have a display of the accuracy of the prediction
of attribute associations. Thus, the techniques for providing such
a display in accordance with the present invention may be applied
to a variety of applications and devices. For instance, the
algorithm(s) of the invention may be applied to the operating
system of a computing device, provided as a separate object on the
device, as part of another object, as a downloadable object from a
server, as a "middle man" between a device or object and the
network, as a distributed object, etc. While exemplary programming
languages, names and examples are chosen herein as representative
of various choices, these languages, names and examples are not
intended to be limiting. One of ordinary skill in the art will
appreciate that there are numerous ways of providing object code
that achieves the same, similar or equivalent parametrization
achieved by the invention.
[0070] The various techniques described herein may be implemented
in connection with hardware or software or, where appropriate, with
a combination of both. Thus, the methods and apparatus of the
present invention, or certain aspects or portions thereof, may take
the form of program code (i.e., instructions) embodied in tangible
media, such as floppy diskettes, CD-ROMs, hard drives, or any other
machine-readable storage medium, wherein, when the program code is
loaded into and executed by a machine, such as a computer, the
machine becomes an apparatus for practicing the invention. In the
case of program code execution on programmable computers, the
computing device will generally include a processor, a storage
medium readable by the processor (including volatile and
non-volatile memory and/or storage elements), at least one input
device, and at least one output device. One or more programs that
may utilize the techniques of the present invention, e.g., through
the use of a data processing API or the like, are preferably
implemented in a high level procedural or object oriented
programming language to communicate with a computer system.
However, the program(s) can be implemented in assembly or machine
language, if desired. In any case, the language may be a compiled
or interpreted language, and combined with hardware
implementations.
[0071] The methods and apparatus of the present invention may also
be practiced via communications embodied in the form of program
code that is transmitted over some transmission medium, such as
over electrical wiring or cabling, through fiber optics, or via any
other form of transmission, wherein, when the program code is
received and loaded into and executed by a machine, such as an
EPROM, a gate array, a programmable logic device (PLD), a client
computer, a video recorder or the like, or a receiving machine
having the signal processing capabilities as described in exemplary
embodiments above becomes an apparatus for practicing the
invention. When implemented on a general-purpose processor, the
program code combines with the processor to provide a unique
apparatus that operates to invoke the functionality of the present
invention. Additionally, any storage techniques used in connection
with the present invention may invariably be a combination of
hardware and software.
[0072] While the present invention has been described in connection
with the preferred embodiments of the various figures, it is to be
understood that other similar embodiments may be used or
modifications and additions may be made to the described embodiment
for performing the same function of the present invention without
deviating therefrom. For example, while exemplary network
environments of the invention are described in the context of a
networked environment, such as a peer to peer networked
environment, one skilled in the art will recognize that the present
invention is not limited thereto, and that the methods, as
described in the present application may apply to any computing
device or environment, such as a gaming console, handheld computer,
portable computer, etc., whether wired or wireless, and may be
applied to any number of such computing devices connected via a
communications network, and interacting across the network.
Furthermore, it should be emphasized that a variety of computer
platforms, including handheld device operating systems and other
application specific operating systems are contemplated, especially
as the number of wireless networked devices continues to
proliferate. Still further, the present invention may be
implemented in or across a plurality of processing chips or
devices, and storage may similarly be effected across a plurality
of devices. Therefore, the present invention should not be limited
to any single embodiment, but rather should be construed in breadth
and scope in accordance with the appended claims.
* * * * *