π§ Daily Study Log [2025-10-11]
Attended lectures on regression modeling, refined crawling logic for the ongoing Social Sentiment Monitor project, and deepened understanding of regression trees through hands-on exercises.
π Class β Big Data Analysis & AI Modeling
- Studied Regression and Regression Trees in theory and practice.
- Learned how linear regression models predict continuous outcomes and how decision treeβbased regression splits data to minimize variance.
- Performed exercises comparing Linear Regression, Ridge/Lasso, and Regression Trees using real datasets.
- Key Takeaway:
- Linear models capture global relationships, while tree-based models excel at modeling nonlinear patterns and interactions.
π» Personal Project β Social Sentiment Monitor
- Continued work on the Social Sentiment Monitor project.
- Improved the news crawling module to handle multiple keywords and dynamic pagination more efficiently.
- Verified API responses and ensured stable data collection from Naver Economic News.
- Next Step:
- Implement sentiment labeling using KoBERT/KcELECTRA.
- Begin preliminary EDA on crawled data to identify tone patterns across media outlets.
β
TL;DR
π Learned regression & regression trees through practical class exercises
π Enhanced crawling logic in Social Sentiment Monitor project
π Prepared for next phase: sentiment labeling & visualization