1: \begin{abstract}

2:  \noindent 

3:    \textit{Submodular continuous functions} are a category of

4:     (generally) non-convex/non-concave functions with a wide

5:    spectrum of applications. We characterize these functions and

6:    demonstrate that they can be maximized efficiently with

7:    approximation guarantees.

8:    Specifically, 

9:    i) We introduce  the \NEWDR\ property that gives a unified characterization of submodularity for all  set, integer-lattice and continuous functions; 

10:    ii) for maximizing monotone {DR-submodular} continuous functions under general  down-closed convex constraints, we propose a

11:    \algname{Frank-Wolfe} variant with $(1-1/e)$ approximation guarantee,

12:    and sub-linear convergence rate; 

13:    iii) for maximizing general \textit{non-monotone}

14:    submodular continuous functions subject to box constraints, we propose a

15:    \algname{DoubleGreedy} algorithm with

16:    $1/3$ approximation guarantee. 

17: %

18:    Submodular continuous functions naturally find

19:    applications in various real-world settings, including influence and revenue

20:    maximization with continuous assignments, sensor energy management,

21: multi-resolution data summarization, 

22:    facility location, 

23:     etc.  Experimental

24:    results show that the proposed algorithms efficiently generate

25:    superior solutions  compared to

26:    baseline algorithms.

27: \end{abstract}

28: