LdaHoffman.this

  1. this()
  2. this(size_t K, size_t W, size_t D, F alpha, F eta, F tau0, F kappa, F eps, TaskPool tp)
    struct LdaHoffman(F)
    this
    (
    size_t K
    ,
    size_t W
    ,
    size_t D
    ,,
    F eta
    ,,,
    F eps = 1e-5
    ,
    TaskPool tp = taskPool()
    )
    if (
    isFloatingPoint!F
    )

Parameters

K size_t

theme count

W size_t

dictionary size

D size_t

approximate total number of documents in a collection.

alpha F

Dirichlet document-topic prior (0.1)

eta F

Dirichlet word-topic prior (0.1)

tau0 F

tau0 ≧ 0 slows down the early iterations of the algorithm.

kappa F

kappa belongs to $(LPAREN)0.5, 1], controls the rate at which old values of lambda are forgotten.

eps F

Stop iterations if ||lambda - lambda'||_l1 < s * eps, where s is a documents count in a batch.

tp TaskPool

task pool

Meta