log_ratio_sample: Log-Ratio Sampling

In activetext/activeR: a semi-supervised active learning algorithm for text classification.

Log-Ratio Sampling

Description

Log-Ratio Sampling

Usage

log_ratio_sample(docs, out, dfm, mu, tau, max_query, edge, regions)

Arguments

docs	[dataframe] Documents with log ratio and cumulative sum columns.
out	[list] List of output objects from 'EM()' function.
dfm	[quanteda dfm] Quanteda dfm object.
mu	Error acceptance rate for first region.
max_query	[integer] Number of documents to be queried.
edge	[logical] Whether we sample from documents that have very high/low posterior
eta	Error acceptance rate for second region.

Value

Sampled documents.

activetext/activeR documentation built on May 31, 2024, 10:21 a.m.