Bayesian Methods For Machine Learning
This post will be a summary of my understanding from the coursera course Bayesian methods for Machine learning.
Currently, I’m working as a PhD student in Florian Jug’s lab. The broad theme of the lab is to apply Machine learning techniques on microscopy data. My project is to do splitting: Split an superimposed image into its constituent channels. Recently published the work in ICCV 23.
Prior to this, I was working as a Research Assistant in National Taiwan University, Taipei under Prof. Hsuan-Tien Lin. I worked on 2 projects:
I received B.Tech+M.Tech in Computer Science in 2015, from the IIT Delhi. I worked as a Data Scientist at Qplum, an online investment advisory firm, from Dec 2015 to Dec 2018.
Vision: 3D Gaze Estimation: (https://www.bmvc2021-virtualconference.com/conference/papers/paper_0643.html). Created a novel architecture which is more robust to variations in magnification levels. Developed a novel method for backward gazes. Achieved state of the art results on Gaze360 and RT-GENE dataset.
Vision: Extreme Rainfall prediction: (https://journals.ametsoc.org/view/journals/aies/1/3/AIES-D-21-0005.1.xml). Using 3D radar data, 2D rain data, altitude data and few others, aim is to predict the rainfall amount in next 3 hours. An image-to-image translation network based approach was taken to solve this problem.
ML in Finance (3 years): Used autoencoder to estimate overall market sentiment which was used to create a market neutral strategy. Used regularized linear regression and other techniques for price prediction.
ML on Biomedical Data (Thesis): Subcellular Regulatory Networks Learning (https://ashesh-0.github.io/Masters/). Jointly learnt gene similarity and gene regulatory network using Markov logic networks. Generated synthetic data and showed the limitations of our work.
This post will be a summary of my understanding from the coursera course Bayesian methods for Machine learning.