• Àüü
  • ÀüÀÚ/Àü±â
  • Åë½Å
  • ÄÄÇ»ÅÍ
´Ý±â

»çÀÌÆ®¸Ê

Loading..

Please wait....

±¹³» ³í¹®Áö

Ȩ Ȩ > ¿¬±¸¹®Çå > ±¹³» ³í¹®Áö > Çѱ¹Á¤º¸°úÇÐȸ ³í¹®Áö > Á¤º¸°úÇÐȸ ÄÄÇ»ÆÃÀÇ ½ÇÁ¦ ³í¹®Áö (KIISE Transactions on Computing Practices)

Á¤º¸°úÇÐȸ ÄÄÇ»ÆÃÀÇ ½ÇÁ¦ ³í¹®Áö (KIISE Transactions on Computing Practices)

Current Result Document :

ÇѱÛÁ¦¸ñ(Korean Title) Áö´ÉÇü µµ½Ã¿¡¼­ ¹Ì¼¼¸ÕÁö ¿¹ÃøÀ» À§ÇÑ ½ÉÃþ ÇнÀ ±â¹ý ÀûÇÕ¼º Æò°¡
¿µ¹®Á¦¸ñ(English Title) Deep Learning Models Conformity Assessment for Particulate Matter Prediction in Smart Cities
ÀúÀÚ(Author) ½ÅÀÍÈñ   ¹®¿ëÇõ   ÀÌ¿ëÁÖ   Ik-Hee Shin   Yong-Hyuk Moon   Yong-Ju Lee  
¿ø¹®¼ö·Ïó(Citation) VOL 25 NO. 12 PP. 0610 ~ 0615 (2019. 12)
Çѱ۳»¿ë
(Korean Abstract)
´ë±âÁß ¹Ì¼¼¸ÕÁö ³óµµ Áõ°¡¿Í ¹Ì¼¼¸ÕÁö°¡ ÀÎü¿¡ ³¢Ä¡´Â Çطοò¿¡ ´ëÇÑ °ü½ÉÀÌ Ä¿Áö¸é¼­ ¹Ì¼¼¸ÕÁö¿¡ ´ëÇÑ ½Åü ³ëÃâÀ» ÁÙÀ̱â À§ÇÑ ¹æ¹ýÀ¸·Î ¹Ì¼¼¸ÕÁö ¿¹ÃøÀÌ ¶°¿À¸£°í ÀÖ´Ù. ¶ÇÇÑ Áö¿ªº° ¹Ì¼¼¸ÕÁö ¼öÄ¡ °üÃø¿¡ ´ëÇÑ ¿ä±¸°¡ Áõ°¡ÇÏ°í ÀÖÁö¸¸ ¹Ì¼¼¸ÕÁö °üÃø¼Ò°¡ ºÎÁ·ÇØ ¿ä±¸¸¦ ¸¸Á·Çϱ⠾î·Æ´Ù. º» ³í¹®¿¡¼­´Â ½º¸¶Æ® ½ÃƼ µ¥ÀÌÅ͸¦ È°¿ëÇÏ¿© ¹Ì¼¼¸ÕÁö °üÃø¼Ò°¡ ºÎÁ·ÇÑ ¹®Á¦¸¦ ÇØ°áÇÏ°í ¹Ì¼¼¸ÕÁö ¼öÄ¡¸¦ ¿¹ÃøÇÏ°íÀÚ ÇÑ´Ù. ½º¸¶Æ® ½ÃƼ µ¥ÀÌÅÍ °°Àº ½Ã°è¿­ µ¥ÀÌÅ͸¦ ´Ù·ç´Â µö·¯´× ¸ðµ¨Àº ÀÔ·Â µ¥ÀÌÅÍÀÇ Æ¯¼º ¼ö¿¡ ¿µÇâÀ» ¹Þ±â¶§¹®¿¡ ¹Ì¼¼¸ÕÁö ¿¹Ãø¿¡ ÀûÇÕÇÑ ¸ðµ¨À» ã¾Æ¾ßÇÑ´Ù. µ¥ÀÌÅÍ Æ¯¼ºÀÌ ´Ù¸¥ µÎ ½º¸¶Æ® ½ÃƼ µ¥ÀÌÅÍ¿¡ ´ëÇØ ´ÙÃþ ½Å°æ¸Á, LSTM, CNN-LSTM ¸ðµ¨À» ÇнÀÇÏ¿© RMSE, MAPEÀÇ °ª°ú Ç¥ÁØÆíÂ÷¸¦ ºñ±³ÇÏ¿© ÀûÇÕÇÑ µö·¯´× ¸ðµ¨À» Á¦½ÃÇÑ´Ù. Á¦½ÃÇÑ ¸ðµ¨°ú ±âÁ¸ ¿¬±¸µé¿¡¼­ »ç¿ëµÈ ¸ðµ¨À» ºñ±³ÇÑ °á°ú Á¦½ÃÇÑ ¸ðµ¨ÀÌ ´õ Á¤È®ÇÑ ¿¹ÃøÀ» ¼öÇàÇÏ¿´´Ù.
¿µ¹®³»¿ë
(English Abstract)
An increase of particulate matter (PM) concentration in the atmosphere and interest about the harmful effects of PM on the human body are attracting more attention. PM prediction takes center stage in reducing the body¡¯s exposure to PM. Also, the demand of local fine dust prediction has increased but the number of PM observation stations are inadequateto satisfy the demand. In this paper, we solve this problem and predict PM value by using smart city data. The deep learning models, using time-series data such assmart city data, are affected by the number of features in data. For this reason, finding suitable models for PM prediction is crucial. We train the multilayer perceptron, LSTM, CNN-LSTM model about two kinds of smart city data which have a different number of features. After training, we compute RMSE and MAPE for each network and suggest the suitable deep learning model based on value of RMSE and MAPE. When we compare the suggested model to various models used in other studies, the suggested model is predicted more precisely.
Å°¿öµå(Keyword) ½º¸¶Æ® ½ÃƼ   ¹Ì¼¼¸ÕÁö ¿¹Ãø   ½Ã°è¿­ µ¥ÀÌÅÍ   µö·¯´×   smart city   particulate matter prediction   time series data   deep learning  
ÆÄÀÏ÷ºÎ PDF ´Ù¿î·Îµå