Rahul Dev

I'm building BazaarBhau, an ongoing bachelor’s major project that forecasts vegetable prices using Kalimati market data and outside factors. I combined XGBoost and GRU models in a stacking ensemble, with Ridge Regression as the meta-learner.

For BazaarBhau, I handle missing values, merge datasets with different time granularities, and convert Nepali (BS) dates to Gregorian. I evaluate the forecasts with RMSE, MAE, and MAPE.

In my minor project, I built a deep-learning prototype to classify chest X-rays as normal or pneumonia. I fine-tuned a ResNet101V2 CNN pretrained on ImageNet and added custom layers to reduce overfitting and improve training convergence.

At YBSolutions, I worked as a TechnicalIntern(ProjectHandling), tracking project tasks and documenting requirements and progress. I’m a fourth-year Computer Engineering student at KantipurEngineering College, and I hold a Professional PythonProgramming Certification from BroadwayInfosys.