Grammar induction using bit masking oriented genetic algorithm and comparative analysis

摘要: A background on theory of grammar induction is presented.The effect premature convergence discussed in detail.Proposed a system for inference by utilizing the mask-fill reproduction operators and Boolean based procedure with minimum description length principle.Comparative analysis, discussion observation obtained results are given an effective manner.Statistical tests (F-test post hoc test) conducted. This paper presents bit masking oriented genetic algorithm (BMOGA) context free induction. It takes advantages crossover mutation together two phases to guide search process from ith generation (i+1)th generation. Crossover operations performed generate proportionate amount population each parser has been implemented checks validity rules acceptance or rejection training data positive negative strings language. Experiments conducted collection regular languages. Minimum principle used corpus samples as appropriate experiment. was observed that BMOGA produces successive generations individuals, computes their fitness at step chooses best when reached threshold (termination) condition. As presented approach found handling therefore compared approaches alleviate convergence. The analysis showed performs better other algorithms such as: random offspring approach, dynamic allocation operators, elite mating pool simple algorithm. term success ratio quality measure its value shows effectiveness BMOGA. Statistical indicate superiority over existing implemented.