Á¤º¸°úÇÐȸ³í¹®Áö (Journal of KIISE)
Current Result Document :
ÇѱÛÁ¦¸ñ(Korean Title) |
¼³¸í °¡´ÉÇÑ KOSPI Áõ°¨ ¿¹Ãø µö·¯´× ¸ðµ¨À» À§ÇÑ Layer-wise Relevance Propagation (LRP) ±â¹Ý ±â¼úÀû ÁöÇ¥ ¹× °Å½Ã°æÁ¦ ÁöÇ¥ ¿µÇ⠺м® |
¿µ¹®Á¦¸ñ(English Title) |
Layer-wise Relevance Propagation (LRP) Based Technical and Macroeconomic Indicator Impact Analysis for an Explainable Deep Learning Model to Predict an Increase and Decrease in KOSPI |
ÀúÀÚ(Author) |
±è»ó¿î
½Å¿øö
Sang-Woon Kim
Won-Chul Shin
ÀÌÀçÀÀ
ÇÑÁöÇü
Jae-Eung Lee
Ji-Hyeong Han
|
¿ø¹®¼ö·Ïó(Citation) |
VOL 48 NO. 12 PP. 1289 ~ 1297 (2021. 12) |
Çѱ۳»¿ë (Korean Abstract) |
ÁÖ½Ä ¿¹Ãø ºÐ¾ß¿¡¼ ÀΰøÁö´ÉÀº Á¤È®µµ¸¦ Çâ»ó½ÃÅ°´Â ¹æÇâÀÇ ¿¬±¸°¡ ÁÖ¸¦ ÀÌ·é´Ù. ÇÏÁö¸¸ ±ÝÀ¶ ºÐ¾ß¿¡¼´Â ¸ðµ¨ÀÇ ¼º´É»Ó¸¸ ¾Æ´Ï¶ó ÀÇ»ç°áÁ¤¿¡ ´ëÇÑ ½Å·Ú¼º°ú Åõ¸í¼º, °øÆò¼ºÀÌ º¸ÀåµÇ¾î¾ß ÇÑ´Ù. ÀÌ¿¡ º» ³í¹®¿¡¼´Â ÁÖ½Ä ¿¹Ãø¿¡ ¸¹ÀÌ »ç¿ëµÇ´Â °Å½Ã°æÁ¦ ÁöÇ¥¿Í ±â¼úÀû ÁöÇ¥¸¦ ÀÔ·Â º¯¼ö·Î ¼±Á¤ÇØ ¸ðµ¨À» ÇнÀ½ÃÅ°°í ÀÌ ¸ðµ¨ÀÇ ¼³¸í°¡´É¼ºÀ» ÁÙ ¼ö ÀÖ´Â LRP ±â¹ýÀ» Á¦¾ÈÇÑ´Ù. ¶ÇÇÑ »ç¿ëÀÚ ÀÔÀå¿¡¼ Á÷°üÀûÀ¸·Î ¸ðµ¨ °á°ú¸¦ È°¿ëÇÒ ¼ö ÀÖµµ·Ï KOSPI ÁÖ°¡ Á¾°¡ÀÇ Àü³¯ ´ëºñ Áõ°¨À¸·Î ¹®Á¦ Á¤ÀǸ¦ °£¼ÒÈÇÏ¿´´Ù. Àû¿ë½ÃŲ LRP¸¦ ÅëÇØ ³ª¿Â ºÐ¼®ÀÇ °á°ú°¡ ½ÇÁ¦ À¯ÀǹÌÇÑ °á°úÀÎ °ÍÀ» º¸À̱â À§ÇØ ºñ±³ ½ÇÇèÀ» ÁøÇàÇÏ¿´´Ù. ½ÇÇè °á°ú LRP¸¦ ÅëÇØ ¼±Á¤ÇÑ º¯¼öµé·Î µ¥ÀÌÅ͸¦ ÇнÀÇÑ ¸ðµ¨ÀÌ ±âÁ¸ÀÇ ¸ðµ¨º¸´Ù ¼º´ÉÀÌ ¿ì¼öÇÔÀ» º¸¿´´Ù. ¶ÇÇÑ, °¢ º¯¼öµéÀÌ ¿¹Ãø°ª¿¡ ±àÁ¤Àû ¿µÇâÀ» ÁÖ´Â °æÇ⼺¿¡ ´ëÇØ ºÐ¼®ÇÏ¿© LRPÀÇ ºÐ¼® °á°ú°¡ À¯ÀǹÌÇÔÀ» º¸¿´´Ù. |
¿µ¹®³»¿ë (English Abstract) |
Most of the research on stock prediction using artificial intelligence has focused on improving the accuracy. However, reliability, transparency, and equity of decision-making should be secured in the field of finance. This study proposes a layer-wise relevance propagation (LRP) approach to create an explainable stock prediction deep learning model, which is trained using macroeconomic and technical indicators as the input features. Also, the definition of the problem is simplified by prediction of an increase or decrease in the KOSPI closing price from the previous day instead of prediction of the KOSPI value itself. To show how the proposed method works, experiments are conducted. The results show that the model trained with data by the selected features via LRP is more accurate than the vanilla model. Moreover, we show that LRP results are meaningful by analyzing the tendency of the positive effect of each feature for the prediction results. |
Å°¿öµå(Keyword) |
Áö½Ä ¿Ï¼º
Áö½Ä ±×·¡ÇÁ
µö·¯´×
ÀΰøÁö´É
ÀÓº£µù
Æ®¸®ÇÃ
knowledge completion
knowledge graph
deep learning
artificial intelligence
word embedding
triple
LRP
KOSPI ¿¹Ãø
°Å½Ã°æÁ¦ÁöÇ¥
±â¼úÀû ÁöÇ¥
µö·¯´×
LRP
KOSPI prediction
macroeconomic indicators
technical indicators
deep learning
|
ÆÄÀÏ÷ºÎ |
PDF ´Ù¿î·Îµå
|