Review :
Partitioned Matrix, Inverse of Partitioned Matrix, Partitioned Regression
º» °ÀÇ Áøµµ :
- Online learning :
-> ȸ±ÍºÐ¼®À» ¹Ù¶óº¸´Â µÎ °¡Áö °üÁ¡
a. º¯¼ö°¡ Ãß°¡°¡ µÇ´Â °æ¿ì(°¡·Î·Î ´Ã¾î³ª´Â °æ¿ì)
* °ü·Ã °³³ä : "Control", "Partition", "Inverse of Partitioned Matrix",
"Omitted Variable Bias", "Confounding variables" µî
b. °üÃøÄ¡°¡ Ãß°¡°¡ µÇ´Â °æ¿ì(¼¼·Î·Î ´Ã¾î³ª´Â °æ¿ì)-Online Updating
* °ü·Ã °³³ä : Marginal Learning, Minibatch Updating, Inverse of Sum of Matrices
- Heteroskedasticity ÀÇ °æ¿ì OLS estimatorÀÇ ºÐ»ê °øºÐ»ê Çà·Ä
White's Idea (White's Heteroskedasticity Robust Variance Estimator)
- Weighted Least Square
- Feasible Weighted Least Square
Âü°í ÀÚ·á :
Heteroskedasticity ~ : wooldridge Ch.8, °è·® ³ëÆ®(·ù±Ù°ü, ÀÓ¼±Çý) Àü¹ÝºÎ - Ch.4 |