As one of the important concepts in epidemiology, herd immunity was recommended to control the COVID-19 pandemic. Inspired by this technique, the Coronavirus Herd Immunity Optimiser has recently been introduced, demonstrating promising results in addressing optimisation problems. This particular algorithm has been utilised to address optimisation problems widely; However, there is room for enhancement in its performance by making modifications to its parameters. This paper aims to improve the Coronavirus Herd Immunity Optimisation algorithm to employ it in addressing breast cancer diagnosis problem through feature selection. For this purpose, the algorithm was discretised after the improvements were made. The Opposition-Based Learning approach was applied to balance the exploration and exploitation stages to enhance performance. The resulting algorithm was employed in the diagnosis of breast cancer, and its performance was evaluated on ten benchmark functions. According to the simulation results, it demonstrates superior performance in comparison with other well-known approaches of the similar nature. The results demonstrate that the new approach performs well in diagnosing breast cancer with high accuracy and less computational complexity and can address a variety of real-world optimisation problems.

作为流行病学的重要概念之一，群体免疫被建议用于控制COVID-19大流行。受这项技术的启发，最近推出了冠状病毒群体免疫优化器，在解决优化问题方面展示了有希望的结果。这种特殊的算法已被广泛用于解决优化问题；然而，通过修改其参数，其性能还有增强的空间。本文旨在改进冠状病毒群体免疫优化算法，通过特征选择将其用于解决乳腺癌诊断问题。为此，对算法进行改进后进行离散化。基于反对的学习方法用于平衡探索和利用阶段以提高绩效。由此产生的算法被用于乳腺癌的诊断，并在十个基准函数上评估其性能。根据仿真结果，与其他已知的类似性质的方法相比，它表现出优越的性能。结果表明，新方法在诊断乳腺癌方面表现良好，准确度高，计算复杂度较低，并且可以解决各种现实世界的优化问题。