-Experienced Data Engineer, Data Scientist, and Machine Learning --Engineer with 2 years of professional experience
- Experienced in Cloud Computing building scaleable and fault-tolerant web services (AWS Stack)
- Professional Experience in Natural Language Processing, Predictive Analytics, and Anomaly Detection
- Academic experience in Computer Vision, Malware detection
-
Experience
Data Science Intern @ SJSU
- Built a forecasting model to predict the user traffic on the application using time series models, developed python scripts to automate real-time data retrieval and storage using google scheduler and visualized it using tableau dashboard
- Built a predictive model to improve college ranking, creating a what if dashboard to plug in and analyze impact of KPI on overall rank
-
Projects
Bitcoin Price Prediction using Sentiment and Predictive analytics [Master’s Project] : Built a novel solution to retrieve real time news about Bitcoin from web , generating sentiment using BERT analyzing its effect on price and trading volume of bitcoin to understand purchase patterns, reaction time and key factors effecting the price finally improved the price prediction using hybrid Facebook Prophet model
Conditional Generative Adversarial Networks (CGAN) for Multi-class Malware Detection: Used semi supervised GAN to classify Multi-class Malware families. The training data is a 40 x 1-dimension vector obtained using Word2Vec with N=2 and Window size =5 for top 20 opcodes in each malware family receiving an accuracy of 85.1 %
Real-time Object Detection on low-light/dark Images: Used EnlightenGAN with Contrast Limited Adaptive Histogram Equalization (CLAHE) and Unsharp Mask(USM) to enhance low-light/dark images and Fine-tuned FasterRCNN object detection using transfer learning on 5000 brightened COCO dataset images achieving 15% higher accuracy.