INMODEL= option, 17
MISSING= option, 17
PADJUST= option, 17
PRIORSSEARCH option, 18
PVARS= option, 18
SPLITATDATUM option, 18
SPLITBETWEEN option, 18
PROPORTION= option
ASSESS statement, 20
PRUNE statement, 28
DROPVARS= option, 28
KEEPRULES= option, 28
PRUNED option
INTERACT statement, 26
PRUNEDATA= option
ASSESS statement, 20
PVARS= option
PROC ARBORETUM statement, 18
R
RAM option
PERFORMANCE statement, 27
REDO statement, 28
RESIDUAL option
CODE statement, 22
ROLE= option
SCORE statement, 30
RULES= option
SAVE statement, 29
S
SAVE statement, 28
IMPORTANCE= option, 28
MODEL= option, 28
NODES= option, 29
NODESTAT= option, 29
PATH= option, 29
RULES= option, 29
SEQUENCE= option, 29
SUMMARY= option, 29
SCORE statement, 29
DATA= option, 29
DUMMY option, 29
NODES= option, 30
NOLEAFID option, 30
NOPREDICTION option, 30
OUT= option, 30
OUTFIT= option, 30
ROLE= option, 30
SEARCH statement, 30
SEARCHBINS= option
TRAIN statement, 35
SEQUENCE= option
SAVE statement, 29
SETRULE statement, 30
SMALLRESIDUAL
MISSING= option, 45
SPLIT statement, 31
MISSBRANCH= option, 31
MISSONLY option, 31
VAR= option, 31
SPLITATDATUM
INPUT statement, 26
SPLITATDATUM option
PROC ARBORETUM statement, 18
SPLITBETWEEN
INPUT statement, 26
SPLITBETWEEN option
PROC ARBORETUM statement, 18
SPLITSIZE= option
TRAIN statement, 36
SUBTREE statement, 32
BEST option, 32
SUMMARY= option
SAVE statement, 29
T
TARGET statement, 32
LEVEL= option, 32
ORDER= option, 33
TRAIN statement, 33
ALPHA= option, 33
EXHAUSTIVE= option, 33
INTERVALBINS= option, 33
LEAFSIZE= option, 33
MAXBRANCH= option, 34
MAXDEPTH= option, 34
MAXNEWDEPTH= option, 34
MAXRULES= option, 34
MAXSURROGATES= option, 35
MINCATSIZE= option, 35
MINWORTH= option, 35
NODES= option, 35
SEARCHBINS= option, 35
SPLITSIZE= option, 36
USEVARONCE option, 36
U
UNDO statement, 36
USEINSEARCH
MISSING= option, 45
USEVARONCE option
TRAIN statement, 36
V
VALIDATA= option
ASSESS statement, 20
VAR= option
SPLIT statement, 31
The ASSOC Procedure
The ASSOC Procedure
Overview
Procedure Syntax
PROC ASSOC Statement
CUSTOMER Statement
TARGET Statement
Details
Example
References
Copyright 2000 by SAS Institute Inc., Cary, NC, USA. All rights reserved.
The ASSOC Procedure
Overview
Association discovery is the identification of items that occur together in a given event or record. This
technique is also known as market basket analysis. Online transaction processing systems often provide
the data sources for association discovery. Associations rules are based on frequency counts of the
number of times items occur alone and in combination in the database. The rules are expressed as "if
item A is part of an event then item B is also part of the event X percent of the time." The rules should
not be interpreted as a direct causation but as an association between two or more items. Identifying
creditable associations can help the business technologist make decisions such as when to distribute
coupons, when to put a product on sale, or how to layout items in a store.
Hypothetical association discovery rules include: If a customer buys shoes, then 10% of the time he also
buys socks. A grocery chain may find that 80% of all shoppers are apt to buy a jar of salsa when they
also purchase a bag of tortilla chips. When "do-it-yourselfers" buy latex paint they, also buy rollers 85%
of the time. Forty percent of investors holding an equity index fund will have a growth fund in their
portfolio.
An association rule has a left side (antecedent) and a right side (consequent). Both sides of the rule can
contain more than one item. The confidence factor, level of support, and lift are three important
evaluation criteria of association discovery. The strength of an association is defined by its confidence
factor, which is the percentage of cases in which a consequent appears with a given antecedent. The level
of support is how frequently the combination occurs in the market basket (data base). Lift is equal to the
confidence factor divided by the expected confidence. A creditable rule has a large relative confidence
factor, a relatively large level of support, and a value of lift greater than 1. Rules having a high level of
confidence but little support should be interpreted with caution.
The maximum number of items in an association determines the maximum size of the item set to be
considered. For example, the default of 4 items indicates that up to 4-way associations are performed.
Copyright 2000 by SAS Institute Inc., Cary, NC, USA. All rights reserved.
The ASSOC Procedure
Procedure Syntax
PROC ASSOC <
option(s)>;
CUSTOMER variable-list;
TARGET variable;
Copyright 2000 by SAS Institute Inc., Cary, NC, USA. All rights reserved.