Naive Bayes Model 1

What is

Consider the following text classification problem: training set of n provision of this feature ¹ , text categories, \ (\ {(\ mathbf {T} ^ I, C ^ I) \} _ {I = 1} ^ n \) ²
Online wherein the given text \ (\ mathbf {T} \) , it is determined that the required category.
Naive Bayesian approach is that the operator \ (p (\ mathbf {t }, c) \) maximum \ (c ^ * \) as \ (\ mathbf {t} \ ) categories:
\ [^ C * = \ arg \ max p (\ mathbf
{t}, c) \] wherein,
\ [P (\ mathbf {T}, C) = P (C) \ prod_ {J =. 1} ^ {m} P (t_j | c) \]
and the right \ (the p-(c) \) , \ (the p-(t_j | c) \) is the estimated value instead of training data, the estimated value:

\[ p(c)=\frac{\#\{c^i=c\}}{n} \]

\ [P (t_j | c ^ i = c) = \ frac {\ # \ {c ^ i = c \ quad \ text {and} \ quad t ^ i_j = t_j \}} {\ # \ {c ^ i = c \}} \]
consider the following text classification problem: training set of n provision of this feature ³ , text categories, \ (\ {(\ mathbf {T} ^ I, C ^ I) \} _ {I =. 1 n-^} \) ^{. 4}
now given text feature \ (\ mathbf {T} \) , it is determined that the required category.
Naive Bayesian approach is that the operator \ (p (\ mathbf {t }, c) \) maximum \ (c ^ * \) as \ (\ mathbf {t} \ ) categories:
\ [^ C * = \ arg \ max p (\ mathbf
{t}, c) \] wherein,
\ [P (\ mathbf {T}, C) = P (C) \ prod_ {J =. 1} ^ {m} P (t_j | C) \]
naive Bayesian approach is that the operator \ (p (\ mathbf {t }, c) \) maximum \ (c ^ * \) as \ (\ mathbf {t} \ ) categories:
\ [ \ begin {align} c ^ *
= \ arg \ max p (\ mathbf {t}, c) \ end {align} \] naive Bayesian approach is that the operator\ (p (\ mathbf {t }, c) \) maximum \ (c ^ * \) as \ (\ mathbf {t} \ ) categories:
\ [\ the begin {align = left} C ^ * = \ Arg \ max p (\ mathbf {t} , c) \ end {align} \]

---

1. Wherein m has assumed that each text attributes ↩

2. Superscript indicates the entire record, this record represents a portion of a subscript, i.e. \ (\ mathbf {T} = I ^ [I ^ T_l, T_2 I ^, ..., ^ I T_m] \) ↩

3. Wherein m has assumed that each text attributes ↩

4. Superscript indicates the entire record, this record represents a portion of a subscript, i.e. \ (\ mathbf {T} = I ^ [I ^ T_l, T_2 I ^, ..., ^ I T_m] \) ↩