机器学习5 指数族分布与logistic分布

$$P(x|y = 1,{\eta _1}) = {b_1}(x){e^{({\eta _1}^T{T_1}(x) - {a_1}({\eta _1}))}}$$
$$p(x|y = 0,{\eta _0}) = {b_0}(x){e^{({\eta _0}^T{T_0}(x) - {a_0}({\eta _0}))}}$$
$$\Rightarrow$$
$$p(y = 1|x){\kern 1pt} {\kern 1pt} satisfy{\kern 1pt} {\kern 1pt} {\kern 1pt} \log istic$$
证:
$$p(y = 1|x) = {{p(y = 1)p(x|y = 1)} \over {p(x)}}$$
$$= {{p(y = 1)p(x|y = 1)} \over {p(y = 1)p(x|y = 1) + p(y = 0)p(x|y = 0)}}$$
$$= {{{b_1}(x){e^{({\eta _1}^T{T_1}(x) - {a_1}({\eta _1}))}}} \over {{b_1}(x){e^{({\eta _1}^T{T_1}(x) - {a_1}({\eta _1}))}} + {b_0}(x){e^{({\eta _0}^T{T_0}(x) - {a_0}({\eta _0}))}}}}$$
$$= {1 \over {1 + {{{b_0}\left( x \right)} \over {{b_1}(x)}}{e^{\left( {{\eta _0}^T{T_0}(x) - {\eta _1}^T{T_1}(x)} \right) + {a_1}({\eta _1}) - {a_0}({\eta _0})}}}}$$