Differences
This shows you the differences between two versions of the page.
Both sides previous revisionPrevious revision | Next revisionBoth sides next revision | ||
data_mining:xgboost [2019/05/03 01:10] – phreazer | data_mining:xgboost [2019/05/03 01:13] – [Taylor expansion] phreazer | ||
---|---|---|---|
Line 95: | Line 95: | ||
$$\sum^n_{i=1} [l(y_i, | $$\sum^n_{i=1} [l(y_i, | ||
- | With removed constants | + | With removed constants |
$$\sum^n_{i=1} [g_if_t(x_i) + \frac{1}{2}h_if_t^2(x_i)] + \Omega(f_t)$$ | $$\sum^n_{i=1} [g_if_t(x_i) + \frac{1}{2}h_if_t^2(x_i)] + \Omega(f_t)$$ | ||
So that learning function only influences $g_i$ and $h_i$ while rest stays the same. | So that learning function only influences $g_i$ and $h_i$ while rest stays the same. |