⭐ 欢迎来到虫虫下载站! | 📦 资源下载 📁 资源专辑 ℹ️ 关于我们
⭐ 虫虫下载站

📄 readme.txt

📁 数据挖掘的工具代码(包含fp-tree,appriory
💻 TXT
字号:
readme.txt - December 5th, 2001 - by Laurentiu CristoforWelcome to the distribution of ARMiner, a client-server applicationfor finding association rules in datasets. ARMiner is distributedunder the GNU General Public License, whose text is included in thisdistribution.The arminer.tar.gz file contains the ARMiner distribution (binaries,sources, and documentation) and is organized as follows:copyright.txt          - copyright disclaimergpl.txt                - The GNU General Public Licensereadme.txt             - this filebin/Client/            - ARMiner client binariesbin/Server/            - ARMiner server binariesmanuals/               - ARMiner manualssrc/client/            - ARMiner client specific sourcessrc/server/            - ARMiner server specific sourcessrc/common/            - ARMiner sources needed by both client and                          serverTo compile ARMiner, copy src/common in src/client/ and src/server/,then run 'make' in src/server/ and 'make allClient' in src/client/. Youneed Java 1.2 or later and a make utility. If you do not have a makeutility you can figure out how to compile the sources by taking a lookat the contents of the makefile. You can generate a documentation forthe server sources by running command 'javadocsrc/server/*.java'. ARMiner does not run or compile if you have a Javaversion earlier than 1.2.We recommend that you browse through the documentation beforelaunching ARMiner. See the client manual first for an introduction ofthe user interface.For a quick start, first launch the server:java -jar Server.jarand then launch the client application:java -jar Client.jarYou can login as admin with password renimra. If during your experiments you run out of memory then you can rerunARMiner and allocate more memory to the JVM using the -Xmx commandline option, as shown below:java -Xmx256M -jar Server.jarjava -Xmx256M -jar Client.jarFor questions, comments, send email to Laurentiu Cristofor(laur@cs.umb.edu).Changes history===============December 5th, 2001~~~~~~~~~~~~~~~~~~Server 1.0a   - I have fixed an error in the Itemset.add() method              - I have fixed RandomPoissonDistribution to approximate                a Poisson distribution with large mean by a normal                distribution. This fixes the infinite loop that                occurred in the synthetic data generation if one                selected large values for the average pattern size.April 5th, 2001~~~~~~~~~~~~~~~Server 1.0    - I have fixed the HashTree data structure to handle                large numbers of itemsets.              - I have replaced the quicksort algorithm used for                sorting the mining results with a heapsort that is                faster and uses less memory.              - I have changed the maximum number of results sent to                the client in one batch to 5000.              - The server now displays a short message telling how                many rules resulted from the mining.              - I made some minor changes in error messages and                comments.Client 1.0b   - I have added a SaveAll button to the mining results                dialog to enable the user to save all the results and                not only the portion that has been requested from the                server via the Next button. Note however that if there                are tens of thousands of rules it will take some time                until all the results are transferred.              - I have changed the format in which the rules are saved                by adding a rule number.November 30th, 2000~~~~~~~~~~~~~~~~~~~Server 0.9b   - I have fixed an error in the rule generation procedure                which made it generate fewer rules than it should.November 22nd, 2000~~~~~~~~~~~~~~~~~~~Server 0.9a   - the synthetic data generator has been slightly modified;                the current version should be better from a statistical                point of view but it may be hard to see the differenceClient 1.0a   - Mining result dialog can be resized and the table will                resize too allowing better display of rules              - Mined rules are saved in a format that is much more                readable              - Tables no longer allow swapping of columns              - Replaced the user and group management dialogs with a                 secondary version made by Li so that the lists behave                 in a more helpful way now              - Fixed a bug in the user management dialog that prevented                changing passwords              - Several bug fixes in error handling code              - Keyboard accelerators have been now added for all menus                and buttons of main dialogs              - Text for labels, dialog titles, buttons has been changed                to be more explicit              - Buttons and other controls have been resized and replaced                so that dialogs share more of a common look              - Labels no longer use different fonts and colors for                 different dialogs              - Client sources have been cleaned up and reorganized, better                names have been given for classes so that now it is easier                to find one's way around in the client sources              - Debugging code has been commented so it doesn't clutter                the screen anymoreMaintenance logs have been set in all files modified, detailing thechanges made, for more information look at the sources.June 22nd, 2000~~~~~~~~~~~~~~~Server 0.9    - initial releaseClient ?.?    - initial release, no version given

⌨️ 快捷键说明

复制代码 Ctrl + C
搜索代码 Ctrl + F
全屏模式 F11
切换主题 Ctrl + Shift + D
显示快捷键 ?
增大字号 Ctrl + =
减小字号 Ctrl + -