The 2nd Kookmin University AI Big Data Analysis Competition Awards Ceremony Held
- 24.12.19 / 이정민
Kookmin University (President Jung Seung-ryul) Graduate School of Business (Dean Kim Do-hyeon) held the award ceremony for the “2nd Kookmin University AI Big Data Analysis Competition” on Friday, December 13.
This year's competition, which is in its second year, was sponsored by the Korea Association of Machinery Industry and began on October 21 and ended on November 17 with the aim of developing a one-year trade volume prediction algorithm using trade data and external data for five major industrial items. A total of 111 people from 43 teams participated in the competition, which was fierce in order to improve the accuracy of the prediction algorithm. The competition was evaluated in the first round and verified in the second round, and as a result, the top four teams (one grand prize, one first prize, and two second prizes) were selected and the winners were awarded certificates and prizes.
The team that won the grand prize said that they took care not to leak data in order to develop a time series-based prediction algorithm. Data Leakage is a problem that significantly reduces the usability of the model by using data at the time of prediction in the learning process.
The team that won the grand prize said that they took care not to leak data in order to develop a time-series-based prediction algorithm. This competition requires the use of external data to predict 180 data points for one year, so the possibility of such problems is high. Therefore, the team said that they used the Pseudo Labeling technique to prevent data leakage. In addition, the team members were able to achieve a stable score by using the stacking technique to combine multiple algorithms, and they said that their efforts to finish in the lower ranks until the end of the competition led to the award.
“First of all, I would like to thank all the teams that competed fiercely while taking classes during the semester,” said Kim Do-hyeon, Dean of the Graduate School of Management. ”I hope that the experience of solving these real-world problems with AI will be a great source of nourishment, and we will continue to hold competitions to support the growth of core talent in the AI field.”
|
This content is translated from Korean to English using the AI translation service DeepL and may contain translation errors such as jargon/pronouns. If you find any, please send your feedback to kookminpr@kookmin.ac.kr so we can correct them.
|
|
The 2nd Kookmin University AI Big Data Analysis Competition Awards Ceremony Held |
||||
|---|---|---|---|---|
|
2024-12-19
41
Kookmin University (President Jung Seung-ryul) Graduate School of Business (Dean Kim Do-hyeon) held the award ceremony for the “2nd Kookmin University AI Big Data Analysis Competition” on Friday, December 13.
This year's competition, which is in its second year, was sponsored by the Korea Association of Machinery Industry and began on October 21 and ended on November 17 with the aim of developing a one-year trade volume prediction algorithm using trade data and external data for five major industrial items. A total of 111 people from 43 teams participated in the competition, which was fierce in order to improve the accuracy of the prediction algorithm. The competition was evaluated in the first round and verified in the second round, and as a result, the top four teams (one grand prize, one first prize, and two second prizes) were selected and the winners were awarded certificates and prizes.
The team that won the grand prize said that they took care not to leak data in order to develop a time series-based prediction algorithm. Data Leakage is a problem that significantly reduces the usability of the model by using data at the time of prediction in the learning process.
The team that won the grand prize said that they took care not to leak data in order to develop a time-series-based prediction algorithm. This competition requires the use of external data to predict 180 data points for one year, so the possibility of such problems is high. Therefore, the team said that they used the Pseudo Labeling technique to prevent data leakage. In addition, the team members were able to achieve a stable score by using the stacking technique to combine multiple algorithms, and they said that their efforts to finish in the lower ranks until the end of the competition led to the award.
“First of all, I would like to thank all the teams that competed fiercely while taking classes during the semester,” said Kim Do-hyeon, Dean of the Graduate School of Management. ”I hope that the experience of solving these real-world problems with AI will be a great source of nourishment, and we will continue to hold competitions to support the growth of core talent in the AI field.”
|
||||
