Employed deep learning and computer vision to implement the unsupervised learning algorithm, Generative Adversarial Networks (GAN), which we used to generate realistic ultrasound images. The data agumentation allowed imputation of missing data and sharper imaging, facilitating more accurate detection of breast tumors, using a Pytorch framework and NVIDIA GPUs. We currently have a patent pending on the application of this technology.
Assisted several startup companies in developing recommender systems for their respective products. Utilized social media APIs to scrape profiles for users' interests, and worked with the team to flesh out use-cases and streamline design flow.
Built a dynamic time series model to predict energy price trends by evaluating risks in renewable energy infrastructure, with an emphasis on long-term accuracy and adaptability. Used BeautifulSoup to scrape 15 years worth of unstructured data, and Python to build an ARIMA model
Hello, my name is Sandeep Tiwari. I am a recent graduate from UC Berkeley, and I'm currently pursuing my Master's in Data Science from the University of Washington. During my free time, I like to work on personal projects and improve my own skills, as steps to becoming the best version of myself that I can be.
As a data scientist, I hope to be able to utilize my skillset to automate tools and processes in the medical sector. I am also very interested in applying ML and AI in the financial sphere, to create extremely accurate predictive models and analyze economic trends.
My goal is to provide value to society. Data is my pathway towards that goal
Feel free to contact me. I would love to chat! You can reach me at tiwaristiwari97@gmail.com or any of my social links below
