AP-NSGA-II: An Evolutionary Multi-objective Optimization Algorithm Using Average-Point-Based NSGA-II

A Fuzzy Rough Model Using Genetic Algorithm for Accuracy Enhancement

“…The final step is running the Genetic algorithm on Finally, we compare our model with Preselection with niches [16], NSGA-II (Non dominated sorting genetic algorithm II) [17,18], ENORA (Evolutionary Non dominated sorting with Radial slots) [19,20] and AP-NSGA-II (Average-Point dominated sorting genetic algorithm II) [21] algorithms on iris dataset. The result shows that our algorithm is better than the others in term of Accuracy (CR -evaluation) and CRtraining and also in the number of fuzzy sets.…”

“…Our algorithm using Genetic algorithm technique to generate another new fuzzy rule from the initial rules then calculates their accuracy again which will be greater than the old rules before using genetic algorithm. The proposed model is applied on the Iris and Wine datasets and the results compared with other models: Preselection with niches [16], NSGA-II (Non dominated sorting genetic algorithm II) [17,18], ENORA (Evolutionary Non dominated sorting with Radial slots) [19,20], AP-NSGA-II (Average-Point dominated sorting genetic algorithm II) [21], FRFS (Fuzzy Rough Feature Selection) [22], T-FRFS (Threshold Fuzzy Rough Feature Selection) [23] and C-FRFS (C-Means Fuzzy Rough Feature Selection) [24] algorithms in term of number of fuzzy sets (L) and classification rate for evaluating the accuracy of training and test instances (CR) to show its validity.…”

Searching With Direction Awareness: Multi-Objective Genetic Algorithm Based on Angle Quantization and Crowding Distance MOGA-AQCD

“…In this procedure, solutions in less crowded regions have a higher probability of being preserved, and simple 'hill-climbing'is employed to escape from local optima. Also In [5] was introduced a variant in NSGA-II called as AP-NSGA-II with significant changes in selection operator that maintain the diversity by creating some average points. Other improvements in the selection process were introduced by replacing the crowding distance with a set of reference points as NSGA-III in [6], where the selection operator is changed unlike NSGA-II to ensure diversity by using a predefined set of reference points that be able to predefine in a structured way.…”