\documentclass{article}
\oddsidemargin=0cm
\evensidemargin=0cm
\textwidth=14.8 true cm
\textheight=23.2 true cm
\def \x{{\mbox{\boldmath $x$}}}
\def \y{{\mbox{\boldmath $y$}}}
\def \a{{\mbox{\boldmath $a$}}}
\def \d{{\mbox{\boldmath $d$}}}
\def \p{{\mbox{\boldmath $p$}}}
\def \r{{\mbox{\boldmath $r$}}}
\def \t{{\mbox{\boldmath $t$}}}
\def \w{{\mbox{\boldmath $w$}}}
\def \fx{{\mbox{\boldmath $\tilde x$}}}
\def \fy{{\mbox{\boldmath $\tilde y$}}}
\def \hxi{{\mbox{\boldmath $\xi$}}}
\def \heta{{\mbox{\boldmath $\eta$}}}
\def \htau{{\mbox{\boldmath $\tau$}}}
\def \Pos{{\rm Pos}}
\def \Ch{{\rm Ch}}
%\font\euler=EUSM10
\begin{document}

\noindent Liu B., {\em Uncertain Programming}, John Wiley \& Sons, New York, 1999

\vskip 1.5cm
\noindent {\bf Example 2.1} This example is related to maximizing a function over a nonconvex set,
$$\left\{\begin{array}{l}
\max f(\x)=\displaystyle\frac{x_1^2x_2x_3^2}{2x_1^3x_3^2+3x_1^2x_2^2
      +2x_2^2x_3^3+x_1^3x_2^2x_3^2}\\[0.3 cm]
\mbox{subject to:}\\[0.1 cm]
\qquad x_1^2+x_2^2+x_3^2\ge  1\\[0.1 cm]
\qquad x_1^2+x_2^2+x_3^2\le 4\\[0.1 cm]
\qquad x_1,x_2,x_3> 0.
\end{array}\right.$$

\end{document}