Mehul Jain Email and Phone Number

Collaborated with Prof. Deepak Pullugram Reddy in the Department of Electrical Engineering. Conducted an extensive literature review and comprehensive data analysis. Performed feature selection using heatmaps and feature importance algorithms, trend analysis employing line plots and Hodrick-Prescott filter, and seasonality evaluation through box plots and autocorrelation charts. Applied sophisticated data preprocessing methods, including seasonal decomposition and pandas differencing, to effectively eliminate underlying trends and seasonality patterns from datasets, optimizing them for analytical modeling.Developed Recurrent Neural Networks - LSTM accurate power forecasting and time series forecasting

Developed data-driven and strategic models to optimize marketing campaigns and enhance customer acquisition strategies. By creating a Look-Alike Audience Model, boosted acquisition by 15% in just 4 months, enabling clients to target customers in similar segments. Increased ROI by 20% in global marketing efforts by analyzing 1.5 billion behaviors per month and identifying 500 million potential customers

Collaborated with Prof. Paresh Nath Singha Roy in the Department of Geology and Geophysics at IIT Kharagpur . Extracted seismic data using the segy.io library. Implemented Python Patchify to partition the seismic dataset into discrete patches. Developed a U-Net architecture—a powerful encoder-decoder framework—using PyTorch for creating a convolutional neural network achieving an impressive accuracy of 95.68%.