Assessment of genetic algorithms for feature selection in the context of classification tasks. The underlying assumption is that using smaller feature subsets in classification might have a positive impact on both the accuracy and the running time of the classifier. The presented approach is using self-implemented Genetic Algorithms as the search tool through the space of feature subsets and K Nearest Neighbors classifier for evaluation.
Started as a seminar work for the Bio-Inspired Artificial Intelligence course, this work is still in progress. The current version of the paper reporting the findings can be found in the respective folder.