Clustring iris flower data set based on this kaggle data set with genetic algorithm in comparison to Kmeans.
This dataset is widely recognized in pattern recognition literature and comprises three classes, each containing 50 instances representing different types of iris plants. Notably, one class is linearly separable from the other two, while the latter two are not linearly separable from each other.
After conducting several tests using a genetic algorithm, the model achieved an impressive accuracy of 96%. It's important to note that this result surpassed the performance of the KMeans algorithm, which achieved an accuracy of 89.3%.
- Be aware that the result of clustring with genetic algorithm could be different due to the randomness of the algorithm, but the best Ive got was 96%.
- Ive implemented my own selection, crossover and mutation which can be found on main.ipynb