Add Your Heading Text Here
Add Your Heading Text Here
Add Your Heading Text Here
Add Your Heading Text Here
Add Your Heading Text Here
Add Your Heading Text Here
Add Your Heading Text Here

Nika Gorchakova

Project Title

This project uses Machine Learning to extract light curves from observations of exoplanet transits. Exoplanets are planets around stars other than the Sun. When they pass directly between the observer and their host star, a transit is observed, causing the star to dim briefly. Transits may be observed by a wide variety of telescopes from the most advanced space telescopes such as JWST to small privately held telescopes. Ground based observations in particular have many sources of uncertainty in the data. This produces highly heterogenous datasets which require significant cleaning and calibration. Existing algorithmic solutions to image processing require significant manual input and this can pose a challenge, especially for non-expert users. Additionally, the manual input constrains scalability by the availability of people willing to act as data processors. This project will evaluate a variety of existing image processing software and identify opportunities for automation and the implementation of machine learning solutions. Machine learning has the potential to fully automate and accelerate image processing pipelines. By doing so, this project will enable scalable solutions for large-scale image processing in the astronomical domain. This will provide a key capability which will enable large scale public participation, processing of astronomical survey data streams and data mining of archival data.

Project Description