Improved Cosine Similarity Measures of Simplified Neutrosophic Sets for Medical Diagnoses: Suggested Modifications

Due to the significant and precise role of similarity measures in medical diagnosis and pattern recognition, Ye (Artificial Intelligence in Medicine, 2015, https://doi.org/10.1016/j.artmed.2014.12.007) proposed an enhanced version of cosine similarity measures for overcoming drawbacks of other existing measures. His proposed measures included both single and interval neutrosophic cosine similarity measures. The proposed cosine similarity measures have applied in medical diagnosis problems. After a deep study of Ye’s measures of similarity, it is observed that the proposed measures by Ye have unreasonable results in some cases, and this will get the decision-maker into trouble in practical applications, especially in the medical domain. The objective of this research is to make the researchers aware of the logical errors considered by Ye in his suggested measures, as well as for introducing modified measures of similarity which are able to overcome all existing drawbacks.