Twitter Sentiment Study is a difficult task that comprises the various kind of preprocessing phases, including reduction in dimensionality. The reduction in dimensionality ensures minimum computational complexity and improved performance in the classification course. In Twitter data,
each tweet has functionality values that may or may not reflect an individual’s response. As a result, when tweets are signified as feature matrices, many sparse data points are created and possibly overhead and error rates increase in sentiment analysis on Twitter. This paper proposes
a novel kind of algorithm as Artificial Bee Colony and Pigeon Inspired Optimization Hybrid Feature Selection Algorithm. The ABC-PIO combines with the characteristics that ABC can produce various samples, PIO can reach the best value rapidly and Cauchy perturbation strategy can improve optimal
solution. The proposed technique archive Accuracy of 99.01% for Decision tree, 77.34% for Navy Bias and 60.89% Random Forest. The comparative analysis show that the proposed ABC-PIO with Decision tree archive much better results compared to other existing techniques.
A Multi-objective Non-Dominated Sorted Artificial Bee Colony Feature Selection Algorithm for Medical Datasets
