There has been a major and rising interest in India for increasing vaccination rate among peoples to make the nation healthier and safer. In this paper, a new hybrid deep learning model is proposed to predict and target vaccination rates in the less immunized regions. The Rank-Based Multi-Layer Perceptron (R-MLP) hybrid deep learning framework uses the data collected from the recently updated District Level Household Survey-4 (DLHS). R-MLP model predicts and categorizes the percentage of partly immunized vaccination rates as extreme, low and medium ranges. This predicted findings are cross-verified by Deep Soft Cosine Semantic and Ranking SVM based model (DSS-RSM). DSS-RSM model uses the data obtained from the medical practitioners through a location-based social network. The proposed model predicts and extracts patterns with high similarity frequency for identifying vulnerable low immunization regions. It classifies the predicted patterns into two classes such as Class 1 is denoted as high ranked regions and Class 2 is denoted as low ranked regions based on the percentage of pattern matches. Finally, the results from R-MLP and DSS-RSM models are cross-linked together using ensemble model. This model finds the loss values to identify the target regions were health care program need to be conducted for increasing the level of immunization among children’s. The proposed hybrid deep learning models trains and validates using python-based Keras and TensorFlow deep learning libraries. The performance of the proposed hybrid deep learning model is compared with other variant machine learning techniques such as Decision Tree C5.0, Naive Bayes and Linear Regression. This comparative results are evaluated using evaluation measures such as Precision, Recall, Accuracy and F1-Measure. Our results show that the hybrid deep learning system is clearly superior to any other alternative approach.

中文翻译：

---

一种混合深度学习模型，用于预测和确定免疫接种较少的区域，以提高儿童的疫苗接种率

印度对提高各国人民的疫苗接种率以使该国更健康，更安全产生了浓厚的兴趣。在本文中，提出了一种新的混合深度学习模型，以预测和确定免疫程度较低地区的疫苗接种率。基于等级的多层感知器（R-MLP）混合深度学习框架使用从最近更新的地区级家庭调查4（DLHS）收集的数据。R-MLP模型将部分免疫接种率的百分比预测并分类为极端，低端和中等范围。此预测结果已通过Deep Soft Cosine语义和基于排名SVM的模型（DSS-RSM）进行了交叉验证。DSS-RSM模型使用通过基于位置的社交网络从医生那里获得的数据。所提出的模型预测并提取具有高相似频率的模式，以识别脆弱的低免疫区域。它将预测模式分为两类，例如，基于模式匹配的百分比，类别1表示为高排名区域，类别2表示为低排名区域。最后，使用集成模型将R-MLP和DSS-RSM模型的结果交叉链接在一起。该模型发现损失值，以识别需要进行医疗保健计划以提高儿童免疫水平的目标区域。拟议的混合深度学习模型使用基于Python的Keras和TensorFlow深度学习库进行训练和验证。所提出的混合深度学习模型的性能与其他各种机器学习技术（例如决策树C5.0，朴素贝叶斯和线性回归）进行了比较。使用诸如“精度”，“召回率”，“准确性”和“ F1-Measure”之类的评估手段对比较结果进行评估。我们的结果表明，混合深度学习系统明显优于任何其他替代方法。