Classification of objects using fuzzy AI optimized using genetic algorithm You can read the complete research
- conference paper
- thesis - coming soon
Navigating the mysteries of AI 'black boxes' can be challenging. However, my thesis introduces an innovative approach that increases model explainability and surpasses traditional accuracy levels. We focus on classifying rice and raisin types using computer vision techniques on grayscale images to extract essential attributes. These attributes feed into a 7-input, 2-output Fuzzy Inference System (FIS), further optimized using a Genetic Algorithm. The result is a transparent, highly effective classification system that outperforms standard machine learning models in both accuracy and explainability. Our method unlocks the AI 'black box', demonstrating the potential of fuzzy systems and genetic algorithms in advanced AI applications.
The model outperforms by 95% and 87% in Rice and Raisin data sets.
- Matlab 2012 or above
- Run the data generation.m file to generate data from the excel
- Run the Main.m to generate the results in mat file. Change the GA parameter accordingly
- Run the results.m to read the mat file and get plots
Dipin Nair
Nair, D., Cohen, K., Kumar, M. (2023). Classification of Rice Using Genetic Fuzzy Cascading System. In: Dick, S., Kreinovich, V., Lingras, P. (eds) Applications of Fuzzy Techniques. NAFIPS 2022. Lecture Notes in Networks and Systems, vol 500. Springer, Cham. https://doi.org/10.1007/978-3-031-16038-7_17