We study the reconstruction-algorithm effect on positive and negative ghost imaging (GI). By introducing three GI algorithms (logarithmic GI, exponential GI, and trigonometric GI), we perform numerical simulations and experiments with different GI algorithms, which demonstrate that a positive or negative ghost image can be recovered by modulating the monotonicity of the object signal function. Our work broadens the GI reconstruction algorithm and deepens the understanding of positive and negative GI principle.