Social Spider Algorithm

This is the source code repository of Social Spider Algorithm (SSA) [1]. The algorithm is implemented in multiple programming languages.

[1] James J.Q. Yu and Victor O.K. Li, "A Social Spider Algorithm for Global Optimization," Appl. Soft Comput., vol. 30, pp. 614–627, 2015.

How to use

You may want to edit the source code to adapt the algorithm to solve your optimization problems.

C++

class MyProblem : public Problem {
public:
    MyProblem(unsigned int dimension) : Problem(dimension) {
        // Your problem initialization.
    }
    double eval(const std::vector<double>& solution) {
        // Your objective function implementation.
        return 0.0;
    }
};
SSA ssa(new MyProblem(30), 30);
ssa.run(10000, 1.0, 0.7, 0.1);

Python

def problem(x):
    # Your objective function implementation.
    return np.sum(x**2, 1)
SSA(problem).run()

The Python implementation of SSA is vectorized. No loop is conducted on the python-level during each iteration of the algorithm. The objective function shall accept all solutions in the population per invoke.

MATLAB

func = @(x)sum(x.^2, 2);
SSA(func, 30, 100, 10000, 25, 1, 0.7, 0.1);

The MATLAB implementation of SSA is a copy of the Python code. It is highly vectorized. The objective function shall accept all solutions in the population per invoke. The default matrix shape is popsize*dimension.

More implementations TBD

• Java

Difference

All SSA implementations are coded from the same pseudo-code. Programming language specific optimization is adopted. If significantly different results on a same problem are obtained from these implementations, the C++ one shall be regarded as a standard.

License

All source codes are released under MIT License.