Ç¥ÁØÈ­ Âü¿©¾È³»

TTAÀÇ Ç¥ÁØÇöȲ

Ȩ > Ç¥ÁØÈ­ °³¿ä > TTAÀÇ Ç¥ÁØÇöȲ

Ç¥ÁعøÈ£ TTAK.KO-11.0280-Part2 ±¸Ç¥ÁعøÈ£
Á¦°³Á¤ÀÏ 2021-12-08 ÃÑÆäÀÌÁö 31
ÇѱÛÇ¥ÁØ¸í °ËÁõ¿ë µ¥ÀÌÅͼ¼Æ®ÀÇ ¹ë·±½º ±â¹Ý ÀΰøÁö´É ¼ÒÇÁÆ®¿þ¾î ½Å·Ú¼º Æò°¡ ¹æ¹ý - Á¦2ºÎ: À̹ÌÁö ŸÀÔ ¹ë·±½º µ¥ÀÌÅÍ ¼³°è
¿µ¹®Ç¥Áظí A Method for Evaluating the Reliability of Artificial Intelligence Software Based on the Balance of the Validation Dataset - Part 2: Design of Image Type Balanced Data
Çѱ۳»¿ë¿ä¾à ÃÖ±Ù »ê¾÷º° ´Ù¾çÇÑ ¼­ºñ½º¿¡ ÀΰøÁö´ÉÀÌ Æ÷ÇÔµÈ ¼ÒÇÁÆ®¿þ¾î µµÀÔÀÌ È®»êµÇ°í ÀÖÀ¸¸ç, ÀÌ·¯ÇÑ ¼­ºñ½ºÀÇ Ç°ÁúÀ» Á¿ìÇÏ´Â ÇÙ½É ¿äÀÎÀº ¼ÒÇÁÆ®¿þ¾î¿¡ žÀçµÈ ÀΰøÁö´ÉÀÇ Á¤È®¼ºÀÌ´Ù. Á¤È®¼ºÀ» °ËÁõÇÏ´Â ÀϹÝÀûÀÎ ¹æ¹ýÀ¸·Î, ÇнÀ¿ë µ¥ÀÌÅͼ¼Æ® ¿Ü º°µµÀÇ °ËÁõ¿ë µ¥ÀÌÅͼ¼Æ®¸¦ ±¸¼ºÇÏ¿© ÀΰøÁö´É µ¿ÀÛ °á°úÀÇ Ãâ·Â °ª°ú ºñ±³ÇÏ´Â ¹æ½ÄÀ» »ç¿ëÇÑ´Ù. ÀÌ ¶§ °ËÁõ¿ë µ¥ÀÌÅÍÀÇ Ç¥º» °ÝÂ÷·Î ÀÎÇÏ¿© ƯÁ¤ ¿µ¿ªÀÇ µ¥ÀÌÅÍ°¡ Áö³ªÄ¡°Ô ÆíÁßµÇ¾î »ý»êµÇ°Å³ª ȤÀº µ¥ÀÌÅÍ ¼öÁý °úÁ¤¿¡¼­ÀÇ ÇнÀ µ¥ÀÌÅÍ°¡ ¹èÆ÷µÈ ÀΰøÁö´É ¼ÒÇÁÆ®¿þ¾îÀÇ ¿î¿µ µ¥ÀÌÅ͸¦ Á¦´ë·Î ´ëÇ¥ÇÏÁö ¸øÇÏ°í ƯÁ¤ ¿µ¿ªÀÌ °ú´ë ´ëÇ¥µÇ´Â °æ¿ì°¡ ¹ß»ýÇϸé ÀÌ°ÍÀ» µ¥ÀÌÅͼ¼Æ®ÀÇ ¹ë·±½º°¡ ÀûÀýÇÏÁö ¸øÇÏ¿© ¹ß»ýµÈ »ùÇøµ ÆíÇâ(sampling bias), »ùÇøµ ¿À·ù(sampling error)¶ó°í ÇÑ´Ù. ÀÌ »ùÇøµ ÆíÇâÀº ÀΰøÁö´É ¼ÒÇÁÆ®¿þ¾îÀÇ Á¤È®¼º °ËÁõ °á°ú¸¦ ¿Ö°î½ÃÅ°´Â ÁÖ¿ä ¿øÀÎÀ̱⠶§¹®¿¡ Çö½Ç ¼¼°èÀÇ ´Ù¾çÇÑ ½Ã³ª¸®¿À¸¦ ó¸®ÇÒ ¼ö ÀÖ´ÂÁö¸¦ Æò°¡ÇÏ´Â °üÁ¡¿¡¼­´Â ¹ë·±½º°¡ È®º¸µÈ Æò°¡¿ë µ¥ÀÌÅͼ¼Æ®¸¦ È°¿ëÇÏ¿© ½Å·Ú ¼öÁØÀ» Æò°¡ÇÏ´Â °ÍÀÌ ÇÊ¿äÇÏ´Ù.
ÀÌ¿¡ º» Ç¥ÁØ¿¡¼­´Â Æò°¡¿ë ¡®À̹ÌÁö¡¯ ¹ë·±½º µ¥ÀÌÅͼ¼Æ®¸¦ ±¸ÃàÇÏ´Â °úÁ¤¿¡¼­, ÀÔ·Â µ¥ÀÌÅÍÀÇ Á¶°Ç¿¡ ºÎÇյǴ µ¥ÀÌÅͼ¼Æ®ÀÇ ºÐÆ÷¸¦ °í¸£°Ô Æí¼ºÇϱâ À§ÇØ ¡®À̹ÌÁö¡¯ µ¥ÀÌÅÍÀÇ ´Ù¾çÇÑ °üÁ¡ Ư¡À» ¹Ý¿µÇÑ ¹ë·±½º µ¥ÀÌÅÍ ¼³°è ¹æ¹ýÀ» Á¤ÀÇÇÑ´Ù.
¿µ¹®³»¿ë¿ä¾à Recently, the introduction of software including artificial intelligence in various services by industry is spreading, and the key factor that determines the quality of these services is the accuracy of artificial intelligence embedded in the software. As a general method of verifying accuracy, a separate verification dataset is prepared in addition to the training dataset and compared with the output value of the artificial intelligence result. In this case, due to the sample gap of the verification data, the data in a specific area is produced due to excessive concentration, or the training data in the data collection process is not properly representative of the operational data of the distributed artificial intelligence software and the specific area is over-represented. When occurs, this is called a sampling bias or sampling error caused by an improper balance of the dataset. Since this sampling bias is a major cause of distorting the accuracy verification results of artificial intelligence software, from the perspective of evaluating whether it can handle a variety of real-world scenarios, a balanced evaluation dataset is used to evaluate the confidence level. It is necessary to do. Therefore, in the process of constructing the 'image' balance dataset for evaluation, in this standard, in order to evenly organize the distribution of the dataset that meets the conditions of the input data, we define a balance data design method that reflects the characteristics of various viewpoints of the 'image' data is used.
±¹Á¦Ç¥ÁØ
°ü·ÃÆÄÀÏ TTAK_[1].KO-11.0280-Part2.pdf TTAK_[1].KO-11.0280-Part2.pdf            

ÀÌÀü
µðÁöÅÐ Æ®À© ±â´É ¼º¼÷µµ ¸ðµ¨
´ÙÀ½
»çÀ̹ö-¹°¸® ½Ã½ºÅÛÀÇ ¼­ºñ½º Ư¼º¿¡ µû¸¥ ÀÚ¿ø °ü¸® ÂüÁ¶ ¸ðµ¨