A Bayesian Approach to Multistage Fitting of the Variation of the Skeletal Age Features

Table 1

Models for testing the performance of the Bayesian cut.

𝑚1	𝑦𝑖=𝛽𝑗1+𝛽𝑗2𝑓𝑖+𝜖𝑗𝑖,
	𝐭=(𝑡1,…,𝑡𝑘−1)
𝑚2	𝑦𝑖=𝛽𝑗1+𝛽𝑗2𝑓𝑖+𝛽𝑗3𝑓𝑖^2+𝜖𝑗𝑖,
	𝐭=(𝑡1,…,𝑡𝑘−1)
𝑚3	𝑦𝑖=𝛽𝑗1+𝛽𝑗2𝑓𝑖+𝛽𝑗3𝑓𝑖^2+𝛽𝑗4𝑓𝑖^3+𝜖𝑗𝑖,
	𝐭=(𝑡1,…,𝑡𝑘−1)
𝑚4	𝑦𝑖=𝛽𝑗1+𝛽𝑗2𝑓𝑖+𝛽𝑗3𝑓𝑖^2+𝛽𝑗4𝑓𝑖^3+𝛽𝑗5𝑓𝑖^4+𝜖𝑗𝑖,
	𝐭=(𝑡1,…,𝑡𝑘−1)
𝑚5	𝑦𝑖=𝛽𝑗1+𝛽𝑗2𝑓𝑖+𝛽𝑗3𝑓𝑖^2+𝛽𝑗4𝑓𝑖^3+𝛽𝑗5𝑓𝑖^4+𝛽𝑗6𝑓𝑖^5+𝜖𝑗𝑖,
	𝐭=(𝑡1,…,𝑡𝑘−1)