“…Multiple Correspondence Analysis (MCA) has been applied and proven effective to the research areas ranging from feature selection [236], discretization [237], video semantic concept detection [110-115, 118, 181], to data pruning [117]. In this paper, our previous work [236] is integrated as a preprocessing step, which has been demonstrated to outperforms other existing feature selection methods, such as information gain (IG), Chi-Square measure (CHI), etc., in terms of classification accuracy.…”