{"created":"2023-06-20T13:20:44.979147+00:00","id":788,"links":{},"metadata":{"_buckets":{"deposit":"bb6294a1-9a05-47d3-8f33-acc260c5eff3"},"_deposit":{"created_by":1,"id":"788","owners":[1],"pid":{"revision_id":0,"type":"depid","value":"788"},"status":"published"},"_oai":{"id":"oai:ir.soken.ac.jp:00000788","sets":["2:429:17"]},"author_link":["0","0","0"],"item_1_creator_2":{"attribute_name":"著者名","attribute_type":"creator","attribute_value_mlt":[{"creatorNames":[{"creatorName":"白石, 友一"}],"nameIdentifiers":[{"nameIdentifier":"0","nameIdentifierScheme":"WEKO"}]}]},"item_1_creator_3":{"attribute_name":"フリガナ","attribute_type":"creator","attribute_value_mlt":[{"creatorNames":[{"creatorName":"シライシ, ユウイチ"}],"nameIdentifiers":[{"nameIdentifier":"0","nameIdentifierScheme":"WEKO"}]}]},"item_1_date_granted_11":{"attribute_name":"学位授与年月日","attribute_value_mlt":[{"subitem_dategranted":"2008-03-19"}]},"item_1_degree_grantor_5":{"attribute_name":"学位授与機関","attribute_value_mlt":[{"subitem_degreegrantor":[{"subitem_degreegrantor_name":"総合研究大学院大学"}]}]},"item_1_degree_name_6":{"attribute_name":"学位名","attribute_value_mlt":[{"subitem_degreename":"博士（統計科学）"}]},"item_1_description_12":{"attribute_name":"要旨","attribute_value_mlt":[{"subitem_description":" The three dimensional parallel coordinate plot (3-D PCP) is a visualization method
to detect hidden information in data by using human spatial perception. The 3-D PCP
proposed in this dissertation can display several characteristics of multiple variables
simultaneously. First, it can show all the information of each observation at a glance.
Second, it can make some non-linear structures in data clear. Finally, it is useful to
find piecewise linear relationships of variables and their conditions.
The basic idea of the 3-D PCP has already appeared in the parallel coordinate plot.
The parallel coordinate plot can visualize multi dimensional data on a two dimensional
plane. Coordinates of all variables are set in parallel. In the standard form of parallel
coordinate plot the bottom position of each axis corresponds to the minimum value of
each variable, and the top to the maximum value. One observation corresponds to one
set of connected lines. Parallel coordinate plot can shows the characteristic between
two adjoining axes of variables directly. If two variables have a correlation coefficient
of 1, lines expressing observations are located horizontally. If two variables have
correlation coefficient -l, lines of them cross in one point in the middle between two
axes.
However, relations between two variables whose positions are apart more than two
axes are not clearly shown immediately. Another serious problem of static parallel
coordinate plot is that it is not easy to distinguish one observation from another when
the number of observations is large. To solve these problems, several interactive
techniques have been developed including highlighting by brushing operations. The
3-D PCP can show the same effect as the highlighting by brushing operation in a
parallel coordinate plot by extending it into 3-dimensional space. We choose one
variable as a reference variable, usually a response variable. The 3-D PCP places
connected lines expressing observations in 3-dimensional spaces by sorting them
according to the values of the reference variable. This observation-wise 3-D PCP
representation is useful for illustrating the characteristics of observations such as
outliers.
The 3-D PCP has another representation in which values of observations on each
variable are connected by lines. It is called variable-wise connection representation
and is useful to see relations between the reference variable and other variables. For
example, the connected lines expressing the variable which has strong linear
relationships with the reference variable are located around a straight line expressing
the reference variable. It is well known that the scatterplot matrix can show
relationships between variables clearly. However, if the number of variables is large,
the single scatterplot elements become too small to be seen properly. The 3-D PCP can
show more variables than a scatterplot matrix. It is sometimes more suitable to show
characteristics of data simultaneously than using a static representation of scatterplot
matrlx.
We note that the 3-D PCP can detect particular non-linear relations, i.e. interaction
by two variables, through observation-wise connection representation. This
relationship is detected by the special pattern of the angles of connected lines
expressing observations. As explained earlier, a correlation coefficient near 1 or -1
between two variables whose axes are adjacent produces parallel or crossing patterns
in parallel coordinate plots. Similar patterns can be detected in 3-D PCP. If we find the
change of such patterns at specific values of the reference variable, we conclude that
the structures of data are different for each region of the reference variables. In such
cases, it is natural to divide the data into several groups in which the structures have
no more changes. We propose to draw many 3-D PCPs corresponding to the groups
simultaneously and use a lattice layout, which places many graphs on a grid. We show
that they are useful to identify the interaction between two variables by analyzing
simulation and real data.
We realize our 3-D PCP by using the Java language. Java has several advantages for
implementing modern data visualization methods. It is a pure object oriented
programming language and has well-designed standard graphics libraries which are
useful to realize 2-D and 3-D graphics and interactive graphical user interfaces. These
libraries can work as useful components of statistical graphics and be in incorporated
by using so-called design patterns. Design patterns are suggested solutions to common
problems often appearing in object-oriented software development. Our
implementation is based on several design patterns for generality and reusability. Our
software enables us to analyze data by utilizing advanced interactive operations given
by Java. We show that the 3-D PCP and the software are expected to lead to new
achievements in the field of data visualization.
This dissertation is set out as following. Chapter 1 surveys issues of information
visualization and basic statistical graphics used in multivariate data visualization
such as scatterplot, scatterplot matrix, 3-D scatterplot and parallel coordinate plot. It
also discusses dynamic techniques of data visualization and existing software products
for data visualization such as Mondrian, ParallAX, and GGobi. In Chapter 2, 2-D
parallel coordinate plots are discussed. Important issues at visual data analysis with
parallel coordinate pIot are considered. We introduce several existing works for
extending parallel coordinate plot into 3'dimensional spaces. In Chapter 3, we discuss
our extension of parallel coordinate plot into 3-dimensional space, and several of its
characteristics. We show usefulness of lattice layout to display several 3-D PCP at a
time in Chapter 4. Chapter 5 analyzes three data sets by using our 3-D PCP. In
Chapter 6, we explain details of our software design. Finally, concluding remarks are
given in Chapter 7.
","subitem_description_type":"Other"}]},"item_1_description_7":{"attribute_name":"学位記番号","attribute_value_mlt":[{"subitem_description":"総研大甲第1149号","subitem_description_type":"Other"}]},"item_1_select_14":{"attribute_name":"所蔵","attribute_value_mlt":[{"subitem_select_item":"有"}]},"item_1_select_8":{"attribute_name":"研究科","attribute_value_mlt":[{"subitem_select_item":"複合科学研究科"}]},"item_1_select_9":{"attribute_name":"専攻","attribute_value_mlt":[{"subitem_select_item":"15 統計科学専攻"}]},"item_1_text_10":{"attribute_name":"学位授与年度","attribute_value_mlt":[{"subitem_text_value":"2007"}]},"item_creator":{"attribute_name":"著者","attribute_type":"creator","attribute_value_mlt":[{"creatorNames":[{"creatorName":"SHIRAISHI, Yuichi","creatorNameLang":"en"}],"nameIdentifiers":[{"nameIdentifier":"0","nameIdentifierScheme":"WEKO"}]}]},"item_files":{"attribute_name":"ファイル情報","attribute_type":"file","attribute_value_mlt":[{"accessrole":"open_date","date":[{"dateType":"Available","dateValue":"2016-02-17"}],"displaytype":"simple","filename":"甲1149_要旨.pdf","filesize":[{"value":"360.6 kB"}],"format":"application/pdf","licensetype":"license_11","mimetype":"application/pdf","url":{"label":"要旨・審査要旨","url":"https://ir.soken.ac.jp/record/788/files/甲1149_要旨.pdf"},"version_id":"13d76150-bae9-4c8b-975e-ddb02538ca2d"}]},"item_language":{"attribute_name":"言語","attribute_value_mlt":[{"subitem_language":"eng"}]},"item_resource_type":{"attribute_name":"資源タイプ","attribute_value_mlt":[{"resourcetype":"thesis","resourceuri":"http://purl.org/coar/resource_type/c_46ec"}]},"item_title":"Game-theoretical and statistical study on combination of binary classifiers for multi-class classification","item_titles":{"attribute_name":"タイトル","attribute_value_mlt":[{"subitem_title":"Game-theoretical and statistical study on combination of binary classifiers for multi-class classification"},{"subitem_title":"Game-theoretical and statistical study on combination of binary classifiers for multi-class classification","subitem_title_language":"en"}]},"item_type_id":"1","owner":"1","path":["17"],"pubdate":{"attribute_name":"公開日","attribute_value":"2010-02-22"},"publish_date":"2010-02-22","publish_status":"0","recid":"788","relation_version_is_last":true,"title":["Game-theoretical and statistical study on combination of binary classifiers for multi-class classification"],"weko_creator_id":"1","weko_shared_id":-1},"updated":"2023-06-20T16:11:38.862651+00:00"}