I am working as a Software Engineer IV (Machine Learning) at CTO Org of Juniper Networks, on Resource Prediction using time-series models. In my free time, I write blogs on datasciencepreparation.com. In the past, I have worked as Visiting Researcher at Brown University (Rhode Island Hospital), working on Medical Image Segmentation (Brain Lesions) project. I graduated from the University of Massachusetts, Amherst with Computer Science Major in May 2018, and worked as Machine Learning Engineer(R&D) at Quantiphi Inc for a year. My professional experience includes experimenting with deep learning architectures for Vision and NLP, Kernels, and optimization algorithms.
The following are my most relevant qualifications and accomplishments.
Other Positions:
Languages/Tools/Libraries: Python, Java, SQL,Django, Anaconda, NLTK, Scikit-Learn, Keras, Tensorflow, pytorch, AWS, GCP, D3, Bokeh, Ubuntu, Windows, MATLAB
MS in Computer Science (Cgpa: 3.77)
Courses Taken: Computer Vision, Natural Language Processing, Advanced Machine Learning, Deep Learning, Intelligent Visual Computing, Algorithms, Database
B.Tech in Computer Science (Cgpa: 7.5)
Courses Taken: Calculus, Applied Linear Algebra, Applied Numerical Methods, Software Engineering, Data Structure & Algorithms, Database, Distributed Systems.
This project is a collaborative effort between Lexalytics and Prof. Andrew McCallum's lab at UMass Amherst.
Concept/Theme Roll up:
1. Extracted important keywords from data files (bigrams, trigrams, etc.) and implemented different representation methods after extracting word embeddings from WordNet. such as averaging embeddings, training Network on own dataset and obtaining embeddings to save context.
2. Experimented with the various approach of clustering the same themes together using 2 ways: first, figuring out the best way of representation. second, improving clustering method.
3. Implemented a research paper based on training a feature-rich transformational network, and use the obtained embeddings for clustering.
Skills: Python, Machine Learning, javascript, Visualization, Neural Networks, Scikit-learn, pytorch
Machine Learning Intern Autodesk
May 2017 - Sept 2017
1. Extracted data from architecture models and analyzing it using TSNE, and various unsupervised clustering algorithms such as Gaussian and density-based spatial clustering, for pattern recognition.
2. Experimented with different machine learning models to detect the type of building and its components(room type) using fully-connected neural networks, support vector machine, XGBoost etc.
3. Created a recommendation system for user, suggesting the next steps based on their current building/component Type, via modeling a probabilistic graph and traversing using Bayesian approach.
Skills: Python, Machine Learning, Probability, Statistics, Neural Networks, Keras
1. Worked on CTEPH Classification through CT Scans of Heart and Lungs.
2. Created a real-time object detection and tracking model for custom categories using YOLO Architecture.
3. Created a Medical Image segmentation API using U-Net and SegCaps Architecture for 3D Brain CT Scans.
4. I experimented with traditional vision algorithms and deep learning architectures for real-time object detection and segmentation.
Skills: Python, Google Cloud, Tensorflow, Keras, Deep Learning, Reinforcement Learning.
I am working as Visiting Researcher at RIH 3D Lab of Rhode Island Hospital(Brown University) under the guidance of Professor Derek Merck.
1. Working on 3D Image Segmentation to detect bleed in Head CT Scans.
2. Experiment with deep learning architectures(U-Net, V-Net, DenseNet) for DICOM/NRDD Formatted Image Segmentation.
Appformix Team: Machine Learning, Root Cause Analysis, Time Series
1. Working on Appformix software which enables operators to optimize and visualize infrastructure resources.
2. End to End production of functionality to analyze data for resource optimization.
3. Write, automate, and execute test plans to ensure feature functionality.
Skills: Python, Google Cloud, AWS, Docker, Kubernetes, Machine Learning, Flask.
AI Conference Review Research [June 2019 - Present]:
This is an open-source project associated with the Longitudinal Study on AI Research Conference Paper Rejections proposed by the LatinX in AI Coalition in March of 2019 and conducted in collaboration with the Black in AI Organization and Neural Information Processing Systems Foundation.
Hands-on One-Shot Learning [May 2019- Present]:
Hands-On One-Shot Learning with Python Book starts by explaining the fundamentals of One-Shot learning and helps you understand the concept of learning to learn. You will delve into various algorithms, such as siamese, matching networks, memory augmented neural networks, etc, by implementing them in Pytorch. As you make your way through the book, you will dive into state-of-the-art meta-learning algorithms such as MAML, DAML, and LSTM Meta Learner. In the concluding chapters, you will work through recent trends in one-shot learning such as adversarial meta-learning, task agnostic meta-learning, and meta imitation learning. Link: https://github.com/shruti-jadon/Hands-on-One-Shot-Learning