Errors in a more general sense

Last updated Sep 25, 2022 Edit Source

• Cs ml
• Math stats

For a two-type classification problem: There are two different types of errors: $\varepsilon_1$ and $\varepsilon_2$. You can calculate them like so:

• Type I Error: $\int_{\Omega_2} p(x|\omega_1)dx$
• Type II Error: $\int_{\Omega_1} p(x|\omega_2)dx$

If you call $\omega_2$ the positive group and $\omega_1$ the negative group. Type I becomes false negative and Type II becomes false positive.