| Project Name | Description | Technologies |
|---|---|---|
| Digital Nano Bricks | This is one of my favorite projects. It was aimed at helping the client hyper-personalize their offerings to millennials. We started by gathering data from different social media profiles, preprocessed it, analyzed it to learn about customer behavior. We also built a world listening engine to capture recent trends among millennials and used it to further identification of millennials on social media. I also worked on designing Support Vectore Machine to separate millennials from non-millennials. Using this module, we were able to identify 150,000 millennials. The nature of relationship between two customers and the relationship between customer and recent trends were represented using Neo4j graph. |
Python
Linux
MapReduce
Neo4j
Social Media Analytics
|
| Weenix Operating System | Weenix is a linux-based operating system. Worked towards implementing kernel threads and processes along with synchronization primitives, Virtual File System, Virtual Memory System. | C Linux |
| Multi View Image Generation | The objective of this project was to generate synthetic view from a given 2D image as input. Created three models to achieve the goal, namely Autoencoder, Autoencoder with Pose, Autoencoder with adversarial Training. | |
| Image Classification | Designed a Convolutional Neural Network to classify images present in CIFAR-10 dataset. Utilized Batch Normalization and SoftMax Cross Entropy Loss to improve the accuracy. |
Python
Tensorflow
|