-- 방법1.
WITH
    T(C1, C2, C3, GRP, RNK) AS
        (SELECT 1, 2, 4, 1, 1 FROM DUAL
         UNION ALL
         SELECT 1, 2, 5, 1, 2 FROM DUAL
         UNION ALL
         SELECT 1, 2, 4, 1, 3 FROM DUAL
         UNION ALL
         SELECT 2, 3, 7, 2, 1 FROM DUAL
         UNION ALL
         SELECT 2, 3, 7, 2, 2 FROM DUAL)
SELECT *
  FROM (
           SELECT C1
                , C2
                , C3
                , GRP
                , RNK
                , MIN(C1) KEEP (DENSE_RANK FIRST ORDER BY RNK) OVER (PARTITION BY GRP) C1_FIRST
                , MIN(C2) KEEP (DENSE_RANK FIRST ORDER BY RNK) OVER (PARTITION BY GRP) C2_FIRST
                , MIN(C3) KEEP (DENSE_RANK FIRST ORDER BY RNK) OVER (PARTITION BY GRP) C3_FIRST
             FROM T
       )
 WHERE    C1 != C1_FIRST
       OR C2 != C2_FIRST
       OR C3 != C3_FIRST;


-- 방법2
WITH
    T(C1, C2, C3, GRP, RNK) AS
        (SELECT 1, 2, 4, 1, 1 FROM DUAL
         UNION ALL
         SELECT 1, 2, 5, 1, 2 FROM DUAL
         UNION ALL
         SELECT 1, 2, 4, 1, 3 FROM DUAL
         UNION ALL
         SELECT 2, 3, 7, 2, 1 FROM DUAL
         UNION ALL
         SELECT 2, 3, 7, 2, 2 FROM DUAL)
SELECT T.*
  FROM T
     , (
           SELECT *
             FROM T
            WHERE RNK = 1
       ) T2
 WHERE T.GRP = T2.GRP
       AND T.RNK != 1
       AND (   T.C1 != T2.C1
            OR T.C2 != T2.C2
            OR T.C3 != T2.C3);

WITH t AS
(
SELECT 1 c1, 2 c2, 4 c3, 1 grp, 1 rnk FROM dual
UNION ALL SELECT 1, 2, 5, 1, 2 FROM dual
UNION ALL SELECT 1, 2, 4, 1, 3 FROM dual
UNION ALL SELECT 2, 3, 7, 2, 1 FROM dual
UNION ALL SELECT 2, 3, 7, 2, 2 FROM dual
)
-- 1. 집합 함수 : MINUS --
SELECT c1, c2, c3, grp FROM t WHERE rnk > 1
 MINUS
SELECT c1, c2, c3, grp FROM t WHERE rnk = 1
;
-- 2.1. Self Join --
SELECT *
  FROM t a
     , t b
 WHERE a.grp = b.grp
   AND a.rnk = 1
   AND b.rnk > 1
   AND ( a.c1 != b.c1 OR
         a.c2 != b.c2 OR
         a.c3 != b.c3 )
;
-- 2.2. Self Join Outer --
SELECT b.*
  FROM t a
 RIGHT OUTER JOIN t b
    ON a.grp = b.grp
   AND a.rnk = 1
   AND a.c1  = b.c1
   AND a.c2  = b.c2
   AND a.c3  = b.c3
 WHERE b.rnk > 1
   AND a.rnk IS NULL
;
-- 3. 계층 쿼리 --
SELECT *
  FROM t a
 WHERE LEVEL = 2
 START WITH rnk = 1
 CONNECT BY PRIOR grp = grp
   AND rnk > 1
   AND ( PRIOR c1 != c1 OR
         PRIOR c2 != c2 OR
         PRIOR c3 != c3 )
   AND LEVEL <= 2
;
-- 4. 분석함수 : FIRST_VALUE --
SELECT *
  FROM (SELECT c1, c2, c3, grp, rnk
             , FIRST_VALUE(c1) OVER(PARTITION BY grp ORDER BY rnk) c1_1
             , FIRST_VALUE(c2) OVER(PARTITION BY grp ORDER BY rnk) c2_1
             , FIRST_VALUE(c3) OVER(PARTITION BY grp ORDER BY rnk) c3_1
          FROM t
        )
 WHERE ( c1 != c1_1 OR
         c2 != c2_1 OR
         c3 != c3_1 )
;
-- 5.1. 서브 쿼리 : NOT IN --
SELECT *
  FROM t
 WHERE rnk > 1
   AND (grp, c1, c2, c3) NOT IN (SELECT grp, c1, c2, c3
                                   FROM t
                                  WHERE rnk = 1
                                 )
;
-- 5.2. 서브 쿼리 : NOT EXISTS --
SELECT *
  FROM t a
 WHERE rnk > 1
   AND NOT EXISTS (SELECT 1
                     FROM t
                    WHERE rnk = 1
                      AND grp = a.grp
                      AND c1  = a.c1
                      AND c2  = a.c2
                      AND c3  = a.c3
                   )
;

wITH
    T(C1, C2, C3, GRP, RNK) AS
        (SELECT 1, 2, 4, 1, 1 FROM DUAL
         UNION ALL
         SELECT 1, 2, 5, 1, 2 FROM DUAL
         UNION ALL
         SELECT 1, 2, 4, 1, 3 FROM DUAL
         UNION ALL
         SELECT 2, 3, 7, 2, 1 FROM DUAL
         UNION ALL
         SELECT 2, 3, 7, 2, 2 FROM DUAL)
SELECT *
FROM (         
SELECT t.*, MIN(c1||','||c2||','||c3) OVER(PARTITION BY grp ORDER BY rnk) AS rank_1 FROM t
) 
WHERE c1||','||c2||','||c3 <> rank_1

위 댓글들을 참조하여 변형하여 사용했습니다. -_-;;

아! Min 이 아니라 fitst_value 를 써야겠군요 . ㅡㅡ;