เอกสารการประชุมวิชาการและเสนอผลงานวิจัย มหาวิทยาลัยทักษิณ ครั้งที่ 19 2552 - page 206

„µ¦ž¦³»¤ª·µ„µ¦Â¨³Áœ°Ÿ¨Šµœª·‹´¥ ¤®µª·š¥µ¨´¥š´„¬·– ‡¦´ÊŠš¸É 19 ž¦³‹Îµže 2552
2
Abstract
The purpose of this research was to predict the development of the opportunity problems
studies. A case studies of computer engineering students using neural network technique. This
conducted from neural network to support by Back propagation applied tools for advice to predict
the development of the opportunity problem studies, the teacher to decide or to specify student’s
problem past for improving the performance of the model of neural network. It had two
experiments. In the first experiment, we used two hundred forty-five records by input forty nodes,
hidden eighty nodes, and display one node. The second experiment, we used two hundred sixty
records by input thirty-five nodes, hidden forty nodes, and display two nodes. The finding are: For
the first experiment, an accuracy value was eighty-three point three five percents, the recall value
was seventy-seven point seven eight percents. And balance value or F-Measure was eighty percent.
And for the second experiment, an accuracy value was eighty-eight point six four percents, the
recall value was fifty-three point four two percents. And balance value or F-Measure was sixty-six
point six seven percents. For finding the efficiency of these two experiments was the second
experiment had accuracy higher than the first experiment. So, the model that suitable for a case
study is the second experiment.
Keywords: Neural Network Back propagation predict
‡Î µœÎµ
ž{ ‹ ‹» ´œ™ µ ´ œš µ Š „ µ ¦ «¹„¬ µ Å —o Á ®È œ‡ ª µ ¤Îµ ‡´ ° Š ¦ ³° µ ‹ µ ¦ ¥r š¸ É ž¦¹„¬ µ Áž}œ
°Š‡rž¦³„°š¸É ε‡´ ˜n°‡ªµ¤Î µÁ¦È‹ °Š™µ´œ°»—¤«¹ „¬µš»„®nŠÄœ„µ¦Ÿ¨·˜´–”·˜Ä®oÁ„·—‡»–£µ¡Å¤n ªnµ‹³Áž}œ
—oµœª·µ„µ¦ „µ¦¡´ •œµ»‡¨·„£µ¡ „µ¦ž¦´˜´ ªÄ®oÁ o µ„´´Š‡¤ „µ¦ªµŠÂŸœ ¨³„µ¦Á˜¦¸ ¥¤˜´ªÁ¡ºÉ°ž¦³„°
°µ¸¡ (¡· ·˜ š°Šž¦³¥¼ ¦, 2547) ˜n ž{‹‹» ´ œ£µ¡Âª—¨o°¤ Ťnªnµ‹³Áž}œ·ÉŠÂª—¨o°¤šµŠ´Š‡¤ Á«¦¬“„·‹ šÎµÄ®o
‡ªµ¤´ ¤¡´ œ›rĄ¨o ·—¦³®ªn µŠ°µ‹µ¦¥rš¸É ž¦¹„¬µ„´œ´ „«¹„¬µ¤¸œo °¥¨Š œ´ „«¹„¬µ‹³ °‡Îµž¦¹ „¬µ®¦º° °‡ªµ¤
nª¥Á®¨º °‹µ„°µ‹µ¦¥r š¸Éž¦¹ „¬µÅ¤n ªnµ‹³Áž} œ—oµœ„µ¦Á¦¸ ¥œ Á¦ºÉ °Šª·µ„µ¦ ®¦º°Á¦ºÉ °Šn ªœ˜´ªÁž} œÁ¦ºÉ°Š‡n°œ o µŠ¥µ„
(®§š´¥ «·¦· ·œ°»—¤-„·‹ , 2544 ) ¨³¡ž{®µÁ¤ºÉ°°µ‹µ¦¥rš¸Éž¦¹„¬µ¤¸‹Îµœªœœ´„«¹„¬µš¸É°¥¼nĜ‡ªµ¤—¼ ¨Á¡·É¤¤µ„ ¹Êœ ‹¹ ŠÅ¤n
¤¸Áª¨µ¡°š¸É‹³¡¼ —‡»¥®¦º°Ä®o‡Îµž¦¹„¬µÂœ³œÎµÂ„n œ´ „«¹„¬µ ®¦º°œ´ „«¹ „¬µÅ¤n¤¸‡ªµ¤°—šœ¡°Äœ„µ¦š¸É‡°¥¡„´
°µ‹µ¦¥r š¸Éž¦¹ „¬µ ž{‹‹»´œœ´Êœ ŗo¤¸ „µ¦œÎµÁš‡œ·‡„µ¦¡¥µ„¦–rޞ¦³¥»„˜rčo ĜŠµœšµŠ›»¦„·‹—oµœ˜nµŠ Ç