The following four assumptions are sufficient, with statements understood for -almost every and every :
- is a strictly positive probability density function on the data space.
- .
- There is one such that for all .
- For each relevant , the map is continuous.
The first two assumptions give . The last two allow the dominated convergence theorem to prove continuity of the normalized posterior density, which is equivalent to continuity in total variation distance.
Articles by others on the same topic
There are currently no matching articles.