This repository stores the record of PaKoML study group meeting. PaKoML is a study group for studying machine learning, consisting of three university students from Seoul National University of Science and Technology and Korea University. It started in the winter of 2021 and has been continued to this day.
Date | Url | About |
---|---|---|
2022.03.13. | https://youtu.be/hNDrAaTnqWg | Kaggle Spaceship dataset (Hanbin Park) Kaggle Iris dataset (Jihwan Hong) Paper An Analysis of Reduced Error Pruning (Jihwan Hong) |
2022.03.27. | https://youtu.be/x-mDZ_yJkXk | Kaggle Digit Recognizer (Minjun Kim, Hanbin Park, Jihwan Hong) Paper Going Deeper With Convolutions (GoogLeNet) (Jihwan Hong) |
2022.04.03 | https://youtu.be/6jMpkj06mUQ | Object Detection, R-CNN (Jihwan Hong) |
2022.05.08 | https://youtu.be/UTf-4nhTWsE | Paper You Only Look Once (YOLOv1) (Jihwan Hong) |
2022.06.26. | https://youtu.be/QH4E088gLqk | Lecture CS224n Lecture 1 (Hanbin Park) Paper Bagging Predictors (Jihwan Hong) |
2022.08.15. | https://youtu.be/6KWHaBU1DDA | Lecture CS224n Lecture 7 (Jihwan Hong), CS224n Assignment 3 (Hanbin Park) |