1: \begin{abstract}

2: 		\gls{ltga} is an effective \gls{ea} to solve complex problems using the linkage information between problem variables. \gls{ltga} performs well in various kinds of  single-task optimization and yields promising results in comparison with the canonical genetic algorithm. However, \gls{ltga} is an unsuitable method for dealing with multi-task optimization problems. On the other hand, \gls{mfo} can simultaneously solve independent optimization problems, which are encoded in a unified representation to take advantage of the process of knowledge transfer. In this paper, we introduce \gls{mfltga} by combining the main features of both \gls{ltga} and \gls{mfo}. \gls{mfltga} is able to tackle multiple optimization tasks at the same time, each task learns the dependency between problem variables from the shared representation. This knowledge serves to determine the high-quality partial solutions for supporting other tasks in exploring the search space. Moreover, \gls{mfltga} speeds up convergence because of knowledge transfer of relevant problems. We demonstrate the effectiveness of the proposed algorithm on two benchmark problems: \glsentrylong{clustp} and \glsentrylong{dtf}. In comparison to \gls{ltga} and existing methods, \gls{mfltga} outperforms in quality of the solution or in computation time.

3: 	\end{abstract}

4: