際際滷

際際滷Share a Scribd company logo
惡悽愀悸 悽愀悸 悋惡悋悋惠 惠惡
Data mining step by step
悋惺惆悋惆
:
忰惆惺 悋悧 悋惡悋惠
惺悋 惠惺惘
悋 悋惆悸 悋惺悋惠 悒悴悋惆 悋惡悋悋惠  悋惠 惡 悋 愃惘惡悸 惺悸  悋惠 悋惡悋悋惠 惠惡 惺悸
悋惘悋惘 惠悽悋
悋惡悋悋惠 惠惡
:
悋悋忰惶 悋惺悋惠 悋悋惶愀悋惺 悋悄 悋 惠悋惠 惺 惘惡愀悋 悽悋  惡悋悋惠 惠忰 惺悸 
 悋悧
悋惡悋悋惠   惠忰
,
悋  惡惘 忰悴  惆 惺 惺悋惠 惺 惡忰惓 惠惠愆 惺悸  惡惡愕悋愀
惡悋悋惠
,
愀惡惺悋

悋悧 悋惓惘 惠惶惡忰 悋悋惶愀悋惺 悋悄 悋 愀惘 悋惡悋悋惠   惠忰 惡 悋惘惡愀 惺悸 悽悋  惠 
惺悸 
惠愕 悋惠 悋惠惠愆
(
knowledge discovery from data( KDD
悋愕惠愆悋 悽愀悋惠  悽愀 惠惺惠惡惘 
悋惡悋悋惠 悋惺惆  悋惺惘
1. Data cleaning (惠惴to remove noise and inconsistent data )悋惠悋惷悸
2. Data integration 惠悋 (where multiple data sources may be
3. combined)
悋 悋惆 悋 悋悋愕悋 悋悋悋愀 惠忰惆惆 悋惡悋悋惠 惡悋 惠惘 悋惠 悋惺悋悴悋惠  惘悋忰 惺惆悸
KDD
4. Data selection (悋悽惠悋惘where data relevant 惺悋悸 to the analysis task are
retrieved from the database)
5. Data transformation (where data are transformed into forms
appropriate for mining by performing summary operations)
6. Data mining (process where intelligent methods are applied to data
patterns 悋惡悋悋惠 )悖悋愀
7. Pattern evaluation 悋愀 惠 (to identify the truly interesting patterns
representing knowledge based on measures)
8. Knowledge presentation (where knowledge representation
techniques are used to present mined knowledge to users.
悋惺悋惆 悋悋愕惠惺悋 惺悸 惡 悋惘 悋惷忰 悋 悋惘惆 悋惠惡  忰惶  悋  悋  
悋惡悋悋惠 悋惺惆 
Data base
Query
悋惡 悋惡悋悋惠  悋悋愕惠惺悋
:
-
悋惺 悋 惡 悋悋愕惠惺悋  悋惘 惷忰  悋 悋惠悋 悋惓悋
Data base Query vs. Data mining Query ....
惠悴悋惘 愕  惺悋惠 惺 悋悋愕惠惺悋
base Data
悋惺悋惆 悋惡悋悋惠 悋惺惆悸 
.1
惡悽悋惆 悋惓悋 悋愕 惠  悋 慍惡悋悧 悋悋 悋惺悋惠 惺 悋悋愕惠惺悋
*
.2
 悋惓惘 惡惡愃 悋愆惠惘悋  悋 悋慍惡悋悧 惺 悋悋愕惠惺悋
1000
悋愆惘 悋悽惘  惆悋惘
*
.3
悋忰惡 悋愆惠惘悋  悋 悋慍惡悋悧 悋愕悋悄 惺 悋悋愕惠惺悋
*
mining Data
悋惡悋悋惠 惺 悋惠惡
.1
悋惠惶
(
classification (
愆悋 惺悋 悋惠 悋悋悧惠悋 悋惡愀悋悋惠  惺 悋悋愕惠惺悋
.
.2
悋惺悋惆 悋 悋惠悴惺
(
clustering (
悋惓 愆惘悋悧 惺悋惆悋惠 惆  悋 悋慍惡悋悧 惺 悋悋愕惠惺悋
.
.3
悋悋惘惠惡悋愀 惺悋悋惠 悋惺惆 悋惠愆悋
(
rules association (
惺 惠慍悋 惡愆 愆惘悋悄悋 惠 悋惠 悋愕惺 惺 悋悋愕惠惺悋
悋忰惡
.
DATA MINING MODELS 悋愕惠悽惆 悋惠悋惠 惡惺惷
悋惡悋悋惠 惠惡 
1- Neural Networks
2- Genetic Algorithms
3- Agent Technology
4- Decision Trees
5- Hybrid Models
6- Statistics
惠惶悋 惠 悋惆 惷愃愀 悋惘惠悋惺 惆惘悋愕悸  惠 悋悽 惡悋悋惠 悴惺悸 惓  悋 悋惠悋 悋惓悋
悋惘悋惘 愆悴惘悸 愀惘悸 惡悖愕惠悽惆悋
Decision Trees
惠惷 悴惆惆悸 悴悋悋惠 悋 悋惆悽  愕惠惘悸 慍悋惠 悋 悋愕惺悸 惺惆惆悸 悴悋悋惠 悋 悋惡悋悋惠 惠惡 惠愀惡悋惠 悋惠惆惠
悋 悋 悋
悋悛惠  愕惡
:
1
-
悋惶惘悸 悋悖惺悋
Banking :
悋惘惷 悽悋愀惘 惠忰
2

悋悋悸
Financial :
悋悋愕 惠惡悋惆  悋愃愆 惠忰惆惆
3
-
悋悋惠惶悋悋惠
Telecommunications :
悽惆悋惠悋 悋慍 悋悋愕惠悽惆悋 惠忰惆惆
4
-
悋惠愕
Marketing :
慍惡 悋惆愃惘悋悸 悋悽惶悋悧惶 惡 悋惺悋悸 悋悴悋惆
5
-
悋惶忰悸 惘惺悋悸 悋惠悋
Care Health and Insurance :
悋惆惺悋 惠忰
6
-
悋愀惡
Medicine :
悋悽惠悸 悖惘悋惷 悋悋悴忰悸 悋愀惡悸 悋惺悋悴悋惠 惠忰惆惆
7
-
悋
Transportation :
悋悋 惡 悋惠慍惺 悴惆 惠忰惆惆
8
-
惡悋惠悴慍悧悸 悋惡惺
Retailing :
悋惠惘悴 悋惺悸 惠惆惘
9
-
悋慍惡 惺悋悋惠 悋惆悋惘悸
Management Relationship Customer :
 悋 悋 惡悋慍惡 悋忰惠悋惴 悋悋慍 悋惺 悋惠悽悋
悸
悴惆惆 慍惡 愕惡 
.
悋悋惺悸 悋悋惓悸 
:
惺 悋悋惆悋惘悸 愕悋惺惆悸 愃惘惷 悋惺惘悋 悋惶悋惺 惶惘 悋惘惷 惡悋悋惠 惺 悋惡悋悋惠 惠惡 惺悸 悋悴惘悋悄 惠
悋惠悽悋 
悽愀惘悋 愆 惺 悋惠惺悋 悋 悋 悋 悋 悋慍惡 惺 悋忰  悋  悋 悋惘悋惘
"
悋 悖 悋惶惘 惺
.
惠惠  惡
 
悋惶惘 惠忰 悋 悋愕惡惡  悖 惡 惓 愃惘   悋愕惠惡惺悋惆 惡 悋惓 悋慍惡悋悧 惺 悋惠惘慍
悽愕悋悧惘 
悋惠愕惆惆 惺惆 惡愕惡 悋惘悋惷 惺 悋悴悸
悋悸 悋悋愕惠愕悋惘 愃悸 悽悋惆 惆惺 忰悋愕惡 惡惘悋悴 悋愕惠悽惆 悋惠愀惡 悋悴 
SQL
 悋  惠惺惘
)
Structured Query Language Server 2000 ) Server 2000
悋惡悋悋惠 悋惺惆悸 悋惆悋惘悸 惴悋 惡悋
System
Management Database
惠忰惆 愕忰 愀惘悋悧 惆  悋惴悸 惡悋悋惠 悋惆悋惘悸  愕悋惺惆悸 惠惶 惠  悋
惓
惠悴惆惆悸 惡悋悋惠 悋惺惆悸 惺 悋忰悋惴 悽悋悋   悋惠 悋惡悋悋惠 悋惆悽悋
data mining.pptx
悋忰惆 惡愕惠 惆悋悧悋 惠 悋 悋愆悴惘悸 悋 惺惆悋 惠悋悋 悋惘悋惘 愆悴惘悸 悋愕惡 惡愕 悋惠悴惺悋惠 惠惓 惠
.
悋
惶悸 悽惠悋惘
悋惺惆悸 惶悋惠 悴惺悸 悋悧悸    悋悽愀惘 惶悸 惺悸
Set Attribute Node
悋愆悋愆悸 悖愕 愕愀 
.
悋悋
悋惷忰悸 愃惘  悋惺惆 悖愕悋悄
(
1
Cluster

2 Cluster 

悋 
)
悋惠悴惺悋惠    惠悴惺  
惺 悋惷愃愀 悽悋  悋悋惺惆   惘悗悸  悋惠悴惺 惆悋悽 悋悴惆悸 悋惆 惠惶 悋惠 悋悋惺惆  悴惺悸
悋惺惆悸 愕悋惘 忰惠悋惠 悋 悋惴惘 悋惺惆悸
path Node
悋愆悋愆悸  悋愕 
.
悋悽愀惘 惶悸 惡悋悽惠悋惘悋
Risk
悋悋 悋惠悴惺 惺 悋惷愃愀
1
Cluster
悋愕悧悸 悋忰悋悋惠 悋悽惠悋惘
bad
悋愆  悋惠悴惺悋惠  悋愕悋愕悋 惺 惠惡
(
3
(
 悋愃悋 悋  悋惠悴惺 悋   悋忰悋悋惠 惺惆惆 悋 惠惷忰 忰惓
39
悋 忰悋悸
12
惘惷 悽悋愀惘 悋惠 忰悋悸
愕悧悸
bad
悋忰惠悋悸 惡愕惡悸
18.31
%

27
悴惆悸 惘惷 悽悋愀惘 悋惠 忰悋悸
good
悋忰惠悋悸 惡愕惡悸
82.68
%

More Related Content

data mining.pptx

  • 1. 惡悽愀悸 悽愀悸 悋惡悋悋惠 惠惡 Data mining step by step 悋惺惆悋惆 : 忰惆惺 悋悧 悋惡悋惠
  • 2. 惺悋 惠惺惘 悋 悋惆悸 悋惺悋惠 悒悴悋惆 悋惡悋悋惠 悋惠 惡 悋 愃惘惡悸 惺悸 悋惠 悋惡悋悋惠 惠惡 惺悸 悋惘悋惘 惠悽悋 悋惡悋悋惠 惠惡 : 悋悋忰惶 悋惺悋惠 悋悋惶愀悋惺 悋悄 悋 惠悋惠 惺 惘惡愀悋 悽悋 惡悋悋惠 惠忰 惺悸 悋悧 悋惡悋悋惠 惠忰 , 悋 惡惘 忰悴 惆 惺 惺悋惠 惺 惡忰惓 惠惠愆 惺悸 惡惡愕悋愀 惡悋悋惠 , 愀惡惺悋 悋悧 悋惓惘 惠惶惡忰 悋悋惶愀悋惺 悋悄 悋 愀惘 悋惡悋悋惠 惠忰 惡 悋惘惡愀 惺悸 悽悋 惠 惺悸 惠愕 悋惠 悋惠惠愆 ( knowledge discovery from data( KDD 悋愕惠愆悋 悽愀悋惠 悽愀 惠惺惠惡惘 悋惡悋悋惠 悋惺惆 悋惺惘
  • 3. 1. Data cleaning (惠惴to remove noise and inconsistent data )悋惠悋惷悸 2. Data integration 惠悋 (where multiple data sources may be 3. combined) 悋 悋惆 悋 悋悋愕悋 悋悋悋愀 惠忰惆惆 悋惡悋悋惠 惡悋 惠惘 悋惠 悋惺悋悴悋惠 惘悋忰 惺惆悸 KDD 4. Data selection (悋悽惠悋惘where data relevant 惺悋悸 to the analysis task are retrieved from the database) 5. Data transformation (where data are transformed into forms appropriate for mining by performing summary operations) 6. Data mining (process where intelligent methods are applied to data patterns 悋惡悋悋惠 )悖悋愀 7. Pattern evaluation 悋愀 惠 (to identify the truly interesting patterns representing knowledge based on measures) 8. Knowledge presentation (where knowledge representation techniques are used to present mined knowledge to users.
  • 4. 悋惺悋惆 悋悋愕惠惺悋 惺悸 惡 悋惘 悋惷忰 悋 悋惘惆 悋惠惡 忰惶 悋 悋 悋惡悋悋惠 悋惺惆 Data base Query 悋惡 悋惡悋悋惠 悋悋愕惠惺悋 : - 悋惺 悋 惡 悋悋愕惠惺悋 悋惘 惷忰 悋 悋惠悋 悋惓悋 Data base Query vs. Data mining Query .... 惠悴悋惘 愕 惺悋惠 惺 悋悋愕惠惺悋 base Data 悋惺悋惆 悋惡悋悋惠 悋惺惆悸 .1 惡悽悋惆 悋惓悋 悋愕 惠 悋 慍惡悋悧 悋悋 悋惺悋惠 惺 悋悋愕惠惺悋 * .2 悋惓惘 惡惡愃 悋愆惠惘悋 悋 悋慍惡悋悧 惺 悋悋愕惠惺悋 1000 悋愆惘 悋悽惘 惆悋惘 * .3 悋忰惡 悋愆惠惘悋 悋 悋慍惡悋悧 悋愕悋悄 惺 悋悋愕惠惺悋 * mining Data 悋惡悋悋惠 惺 悋惠惡 .1 悋惠惶 ( classification ( 愆悋 惺悋 悋惠 悋悋悧惠悋 悋惡愀悋悋惠 惺 悋悋愕惠惺悋 . .2 悋惺悋惆 悋 悋惠悴惺 ( clustering ( 悋惓 愆惘悋悧 惺悋惆悋惠 惆 悋 悋慍惡悋悧 惺 悋悋愕惠惺悋 . .3 悋悋惘惠惡悋愀 惺悋悋惠 悋惺惆 悋惠愆悋 ( rules association ( 惺 惠慍悋 惡愆 愆惘悋悄悋 惠 悋惠 悋愕惺 惺 悋悋愕惠惺悋 悋忰惡 .
  • 5. DATA MINING MODELS 悋愕惠悽惆 悋惠悋惠 惡惺惷 悋惡悋悋惠 惠惡 1- Neural Networks 2- Genetic Algorithms 3- Agent Technology 4- Decision Trees 5- Hybrid Models 6- Statistics
  • 6. 惠惶悋 惠 悋惆 惷愃愀 悋惘惠悋惺 惆惘悋愕悸 惠 悋悽 惡悋悋惠 悴惺悸 惓 悋 悋惠悋 悋惓悋 悋惘悋惘 愆悴惘悸 愀惘悸 惡悖愕惠悽惆悋 Decision Trees
  • 7. 惠惷 悴惆惆悸 悴悋悋惠 悋 悋惆悽 愕惠惘悸 慍悋惠 悋 悋愕惺悸 惺惆惆悸 悴悋悋惠 悋 悋惡悋悋惠 惠惡 惠愀惡悋惠 悋惠惆惠 悋 悋 悋 悋悛惠 愕惡 : 1 - 悋惶惘悸 悋悖惺悋 Banking : 悋惘惷 悽悋愀惘 惠忰 2 悋悋悸 Financial : 悋悋愕 惠惡悋惆 悋愃愆 惠忰惆惆 3 - 悋悋惠惶悋悋惠 Telecommunications : 悽惆悋惠悋 悋慍 悋悋愕惠悽惆悋 惠忰惆惆 4 - 悋惠愕 Marketing : 慍惡 悋惆愃惘悋悸 悋悽惶悋悧惶 惡 悋惺悋悸 悋悴悋惆 5 - 悋惶忰悸 惘惺悋悸 悋惠悋 Care Health and Insurance : 悋惆惺悋 惠忰 6 - 悋愀惡 Medicine : 悋悽惠悸 悖惘悋惷 悋悋悴忰悸 悋愀惡悸 悋惺悋悴悋惠 惠忰惆惆 7 - 悋 Transportation : 悋悋 惡 悋惠慍惺 悴惆 惠忰惆惆 8 - 惡悋惠悴慍悧悸 悋惡惺 Retailing : 悋惠惘悴 悋惺悸 惠惆惘 9 - 悋慍惡 惺悋悋惠 悋惆悋惘悸 Management Relationship Customer : 悋 悋 惡悋慍惡 悋忰惠悋惴 悋悋慍 悋惺 悋惠悽悋 悸 悴惆惆 慍惡 愕惡 .
  • 8. 悋悋惺悸 悋悋惓悸 : 惺 悋悋惆悋惘悸 愕悋惺惆悸 愃惘惷 悋惺惘悋 悋惶悋惺 惶惘 悋惘惷 惡悋悋惠 惺 悋惡悋悋惠 惠惡 惺悸 悋悴惘悋悄 惠 悋惠悽悋 悽愀惘悋 愆 惺 悋惠惺悋 悋 悋 悋 悋 悋慍惡 惺 悋忰 悋 悋 悋惘悋惘 " 悋 悖 悋惶惘 惺 . 惠惠 惡 悋惶惘 惠忰 悋 悋愕惡惡 悖 惡 惓 愃惘 悋愕惠惡惺悋惆 惡 悋惓 悋慍惡悋悧 惺 悋惠惘慍 悽愕悋悧惘 悋惠愕惆惆 惺惆 惡愕惡 悋惘悋惷 惺 悋悴悸 悋悸 悋悋愕惠愕悋惘 愃悸 悽悋惆 惆惺 忰悋愕惡 惡惘悋悴 悋愕惠悽惆 悋惠愀惡 悋悴 SQL 悋 惠惺惘 ) Structured Query Language Server 2000 ) Server 2000 悋惡悋悋惠 悋惺惆悸 悋惆悋惘悸 惴悋 惡悋 System Management Database 惠忰惆 愕忰 愀惘悋悧 惆 悋惴悸 惡悋悋惠 悋惆悋惘悸 愕悋惺惆悸 惠惶 惠 悋 惓 惠悴惆惆悸 惡悋悋惠 悋惺惆悸 惺 悋忰悋惴 悽悋悋 悋惠 悋惡悋悋惠 悋惆悽悋
  • 10. 悋忰惆 惡愕惠 惆悋悧悋 惠 悋 悋愆悴惘悸 悋 惺惆悋 惠悋悋 悋惘悋惘 愆悴惘悸 悋愕惡 惡愕 悋惠悴惺悋惠 惠惓 惠 . 悋 惶悸 悽惠悋惘 悋惺惆悸 惶悋惠 悴惺悸 悋悧悸 悋悽愀惘 惶悸 惺悸 Set Attribute Node 悋愆悋愆悸 悖愕 愕愀 . 悋悋 悋惷忰悸 愃惘 悋惺惆 悖愕悋悄 ( 1 Cluster 2 Cluster 悋 ) 悋惠悴惺悋惠 惠悴惺 惺 悋惷愃愀 悽悋 悋悋惺惆 惘悗悸 悋惠悴惺 惆悋悽 悋悴惆悸 悋惆 惠惶 悋惠 悋悋惺惆 悴惺悸 悋惺惆悸 愕悋惘 忰惠悋惠 悋 悋惴惘 悋惺惆悸 path Node 悋愆悋愆悸 悋愕 . 悋悽愀惘 惶悸 惡悋悽惠悋惘悋 Risk 悋悋 悋惠悴惺 惺 悋惷愃愀 1 Cluster 悋愕悧悸 悋忰悋悋惠 悋悽惠悋惘 bad 悋愆 悋惠悴惺悋惠 悋愕悋愕悋 惺 惠惡 ( 3 ( 悋愃悋 悋 悋惠悴惺 悋 悋忰悋悋惠 惺惆惆 悋 惠惷忰 忰惓 39 悋 忰悋悸 12 惘惷 悽悋愀惘 悋惠 忰悋悸 愕悧悸 bad 悋忰惠悋悸 惡愕惡悸 18.31 % 27 悴惆悸 惘惷 悽悋愀惘 悋惠 忰悋悸 good 悋忰惠悋悸 惡愕惡悸 82.68 %