Autism Spectrum Disorder (ASD) is a disorder of neurodevelopment whose delayed diagnosis has been posing a barrier in alleviating the severity of the conditions of the sufferers. ASD patients experience difficulties with social communication and interaction. They also show restricted and repetitive patterns of behavior. Approximately 62 million people are diagnosed with ASD globally. Males are about 3–4 times more likely to suffer from ASD than females. Statistically, ASD can be detected between the age of one to two, but some cases may remain undetected for a substantial period. It is crucial to detect ASD precisely and at the nascent stage to remediate the disease. The presented paper proposes an algorithm, namely Modified Grasshopper Optimization Algorithm (MGOA), capable of detecting Autism Spectrum Disorder at all stages of life. GOA is a nature-inspired algorithm that has the potential to explore and exploit the search space effectively. Through this paper, we have attempted to overcome the shortcomings of the traditional GOA, resulting in early diagnosis of the disease. The algorithm is used on the three ASD screening datasets targeting different age groups, namely children, adolescents, and adults, are used for numerical experimentation, and the results are contrasted with the state-of-the-art algorithms. The proposed algorithm with the Random Forest classifier predicted ASD with an approximate accuracy of 100% with specificity and sensitivity as 100% at all stages of life.

自闭症谱系障碍（ASD）是一种神经发育障碍，其延迟诊断一直在减轻患者病情的严重性方面构成障碍。ASD患者在社交沟通和互动方面遇到困难。它们还显示出受限和重复的行为模式。全球大约有6200万人被诊断出患有ASD。男性患ASD的可能性是女性的3-4倍。从统计上讲，可以在一到两岁之间检测到ASD，但有些情况可能在相当长的一段时间内仍未被发现。准确且在新生阶段检测ASD以补救该疾病至关重要。本文提出了一种算法，即改进的蚱hopper优化算法（MGOA），它能够在生活的各个阶段检测自闭症谱系障碍。GOA是一种受自然启发的算法，具有有效探索和利用搜索空间的潜力。通过本文，我们试图克服传统GOA的缺点，从而导致对该疾病的早期诊断。该算法用于针对不同年龄组（即儿童，青少年和成人）的三个ASD筛选数据集，用于数值实验，并将结果与​​最新算法进行对比。带有随机森林分类器的拟议算法可预测ASD，其在生命的各个阶段的准确度约为100％，而特异性和敏感性为100％。导致疾病的早期诊断。该算法用于针对不同年龄组（即儿童，青少年和成人）的三个ASD筛选数据集，用于数值实验，并将结果与​​最新算法进行对比。带有随机森林分类器的拟议算法可预测ASD，其在生命的各个阶段的准确度约为100％，而特异性和敏感性为100％。导致疾病的早期诊断。该算法用于针对不同年龄组（即儿童，青少年和成人）的三个ASD筛选数据集，用于数值实验，并将结果与​​最新算法进行对比。带有随机森林分类器的拟议算法可以预测ASD，其在生命的各个阶段的准确度约为100％，而特异性和敏感性为100％。