Siddharth Jain

Provided comprehensive analysis and recommend solutions on content ranking for each genre on ZEE5 platform. Performed Data Enrichment to improve ads targeting to improve ads revenue.

Worked on Research and development of Vertical Federated Learning Environment. Worked on Identity resolution using PII data. Developed personal and business Email addresses classification, used traditional machine learning based NLP approach after feature engineering and Deep Learning approach by fine-tuning BERT model. Created household classification of family and non-family households on TBs of data in Big Data Environment. Worked on address ranking and Classification model to identify current and previous address using online and online attributes. Implement on a prototype to get boundary (latitude and longitude of all corners of any object)of buildings on google map using Image processing.

International opportunity to learn, implementation and Collaboration under highly skilled Data Scientist Mentors. Performed Multi-label classification on imbalanced thermal and RGB images dataset, having 5 classes, created 4 channel images(RGB + Thermal), Trained with focal-loss, ensemble learning using PyTorch and FAST AI. Performed transfer learning on Aerial images (Xview dataset) using different pre-trained model (Dota,Fmow and Imagenet dataset) to check performance of transfer learning on Aerial images for Multi label classification.

Developed complete ERP solution for waste management Data collection automatically. Implemented IoT to make system Automated, Data collection using weighing machine, Data Transfer from Arduino,Raspberry pi to Mobile App using Blue tooth and the store into Firebase cloud. Show data on Django Web interface by collecting from Firebase.

It was a great learning opportunity to implement academic web development knowledge (Django) into industrial product.

Developed home automation using Image processing, Machine Learning and IoT. Provided prototype for smart vertical farming and Smart Lock.

Worked on Email classification(Spam/Not-Spam). Email Tagging based on different category (Banking, Education, Insurance, Jobs). Implemented Different Classification techniques (Naive Bayes, SVM, Logistic regression, Neural Network models).

My first internship to get industrial exposure. Worked on Open Source PyScada (Python Based Scada System).