Prediction of H1B Visa Application Processing Status Using Deep Learning

The H1B visa plays a crucial role for people trying to pursue their career in USA, however the entire task of applying for the visa and getting approval is tedious. People with high technical skills are missing their opportunities to work in their dream company because of the denial of visa. The entire visa application process consumes a significant amount of time and money; repeating it after a visa application if initially denied would result in not only time and financial waste but also the loss of a professional opportunity. Automated H1B visa Application Processing System (AHAPS) proposed in this work is an intelligent software system that can predict whether the H1B visa data provided will be approved or denied, well ahead the actual visa processing time. For this prediction it uses a deep neural network that uses backpropagation with multiple hidden layers to learn the input features dense and thereby increase the accuracy of prediction. If suppose the application data is denied, then there is a feature extractor system that identifies the significant features in the input data that has contributed to the application denial. The proposed system performance is evaluated with Kaggle dataset. The deep neural network based AHAPS approach obtained an accuracy of 98.2% for predicting the status of visa processing.