Y
HN Search
Hacker News Search
new
|
comments
|
top
|
jobs
clouden
searching PlanetScale…
1.
▲
2.
▲
3.
▲
4.
▲
5.
▲
6.
▲
4 ms
·
1.
▲
by
clouden
12y ago
Logistic regression is similar to his example on fitting a line to an exponential trend. The differences is that in logistic regression, the equation would be ln(p)/ln(1 - p) = α + Σßᵢxᵢ where p is in the interval [0, 1] and E is . No
2.
▲
by
clouden
12y ago
A p-value of 0.04 is not very strong evidence of an effect, though it is significant at the 0.05 level. This study says at most that the effect should continue to be studied.