The latent factor analysis (LFA) model has been widely used to uncover latent relationships from high-dimensional sparse (HiDS) matrices. However, the performance of LFA depends largely on the hyper-parameter value used in the model training. Traditional hyper-parameter tuning methods such as grid search suffer from inefficiency and inaccuracy. In recent years, the particle swarm optimization (PSO) algorithm offers an intelligent approach to adaptively adjust the hyper-parameter of LFA. However, the global optimal solution of the hyper-parameter tuning problem is not fixed due to its dynamic decision space. Therefore, it is difficult for PSO to determine the best hyper-parameter for each training iteration. To address this problem, this paper proposes a novel hyper-parameter adaptive adjustment algorithm called dynamic stochastic reorientation PSO (DSR-PSO) that adapts to constantly changing decision spaces. By randomly adjusting the search directions of particles and perturbing the elite particles, the dynamic property of the DSR-PSO can be enhanced, so that the hyper-parameter can be adjusted in real time throughout the model training process. Furthermore, this paper proves the convergence of the DSR-PSO and gives its convergence condition by discussing the distribution of the characteristic roots. Finally, this paper proposes the DSR-PSO-based LFA (DPL) model by incorporating the DSR-PSO-based hyper-parameter adjustment into the LFA to promote its model training, and analyzes its complexity. Experimental results on benchmark datasets show that the proposed DPL surpasses state-of-the-art LFA models in terms of accuracy and efficiency.

In latent factor analysis (LFA), stochastic gradient descent (SGD) is a widely used algorithm to decompose high-dimensional sparse matrices. However, traditional SGD relies on manually tuning the learning rate which is a vital hyper-parameter to influence its optimization. To improve the performance of LFA, particle swarm optimization (PSO) has been employed to automatically adjust the learning rate of SGD. However, the classical PSO algorithm can not track the movement of the optimal hyper-parameter setting during its convergence process and suffers from the hysteresis due to its individual update scheme. To solve this problem, this paper proposes an algorithm called enhanced dynamic gradient particle swarm optimizer (EDG-PSO) which can achieve the dynamic global optimization of the learning rate during the model training process of LFA. In EDG-PSO, the directions of particles are updated with a certain probability through calculating their gradient’s changes. Moreover, the incorporation of the Adam algorithm further improves the search efficiency of PSO. By embedding the proposed hyper-parameter tuning algorithm into the SGD-based LFA, the EGD-PSO-incorporated LFA (EPLFA) model is proposed by this paper. The experimental results show that the proposed EPLFA model has a higher prediction accuracy and a lower computational cost than existing adaptive adjustment LFA models.