Generative Training
The training methods described above are discriminative training methods, because they seek to find a function that discriminates well between the different output values (see discriminative model). For the special case where is a joint probability distribution and the loss function is the negative log likelihood a risk minimization algorithm is said to perform generative training, because can be regarded as a generative model that explains how the data were generated. Generative training algorithms are often simpler and more computationally efficient than discriminative training algorithms. In some cases, the solution can be computed in closed form as in naive Bayes and linear discriminant analysis.
Read more about this topic: Supervised Learning
Famous quotes containing the words generative and/or training:
“The generative energy, which, when we are loose, dissipates and makes us unclean, when we are continent invigorates and inspires us. Chastity is the flowering of man; and what are called Genius, Heroism, Holiness, and the like, are but various fruits which succeed it.”
—Henry David Thoreau (1817–1862)
“Unfortunately, life may sometimes seem unfair to middle children, some of whom feel like an afterthought to a brilliant older sibling and unable to captivate the family’s attention like the darling baby. Yet the middle position offers great training for the real world of lowered expectations, negotiation, and compromise. Middle children who often must break the mold set by an older sibling may thereby learn to challenge family values and seek their own identity.”
—Marianne E. Neifert (20th century)