Interval type-2 fuzzy logic systems optimization with swarm algorithms for data classification

Fuzzy systems based on the interval type-2 fuzzy set have many advantages in processing uncertain data compared with the fuzzy systems based on the type-1 fuzzy set. The design of optimal interval type-2 fuzzy systems is often difficult due to many parameters. The selection and construction of membership functions used to map the crisp inputs to fuzzifier data play an important role and greatly influence the accuracy of the fuzzy system. The paper proposes a hybrid optimization model using swarm optimization algorithms to find the parameters for the membership function of the interval type-2 fuzzy logic system (IT2FLS). For the experiment, the paper uses optimization techniques such as particle swarm optimization (PSO), genetic algorithm (GA), ant colony optimization (ACO) to find the optimal parameter for IT2FLS applied to the classification problem. Experimental results on datasets from the UCI machine learning library and satellite image data show that hybrid optimization models between the optimization algorithm and IT2FLS can help IT2FLS achieve higher accuracy in data classification problems.