研究文章雷竞技app下载苹果版

神经网络模型在HIV / AIDS的应用程序

文摘

艾滋病是一种无法治愈的疾病。超过数百万人HIV阳性。然而,新的药物不仅可以感染的进展缓慢,但也可以抑制病毒,从而恢复人体的免疫功能,允许许多HIVinfected个人正常,无病的生活。许多研究正在进行预测更好的治疗艾滋病患者,如艾滋病毒药物预测、耐药性检测,预测的副作用对某些方案等制度规范的预测是一项具有挑战性的研究。因为所有病人医疗历史上是独一无二的,在patricular药物副作用和过敏,医生不能以同样的方式对待所有病人。是很常见的,如果一组患者的某些sympotms咨询医生,病人可能得到不同的意见潜在疾病的类型。医生judegement在这方面是一个重要的角色。最近的研究表明,计算智能已广泛应用在医学diagonosis来解决复杂的问题通过开发决策支持系统与神经网络算法的应用。神经网络是非常好的地区实践大部分的医疗问题。它有许多算法分类,预测图像处理etc.A适当利用神经网络技术来实现规模很大,卫生服务研究的数据集是在神经网络领域最困难的地区。 It is further complicated due to ill-defined and illstructured factors affecting a functional health status of HIV /AIDS patients. Many of the studies have applied Neural Network technique to classify and predict desired solution or to improve methodological aspects. In this proposed work, we have taken 300 HIV / AIDS patient’s medical history and constructed a model to predict the appropriate regimen specification, which could help the patient to prolong their for maximum years. To construct this model we had been implemented Back Propagation Neural Network algorithm, ART1 Network and Radial Basis Function Network. Back Propagation Neural Network algorithm is used for classification and prediction purpose and also it would work with huge amount of data with large number of iterations. Due to its feed backward nature it could be act as better prediction algorithm. Similary the ART1 Neural Network algorithm has used to classify the patients into two groups active and inactive based on their regimen specification and the Radial Basis Neural Network has also used to prdict the regimen specification. All these three algorithms have used in this work to predict better regimen specification for HIV / AIDS patients.

M。莉莉佛罗伦萨和Dr.P.Balasubramanie

阅读全文下载全文