An Algorithm Combining Random Forest Classification and Fuzzy Comprehensive Evaluation

<p>Random forest algorithm is a common classification method. However, if the weights of many attributes in data set are not same or close to each other, direct use this for training will lead neglect interrelationships between these attributes, and it difficult reflect differences brought by different attributes. Worse, number relatively large, be given very little weight when normalization satisfied, which also information loss. All have negative impact on final result. To solve problems, paper proposes an combining random fuzzy comprehensive evaluation, only take into account correlation training, but retain original maximun. At time, significantly improves accuracy results.</p> <p> </p>