ASPCS
 
Back to Volume
Paper: Identifying Exoplanets from Kepler Light Curves: a Partial Replication of a Deep Learning Study by Shallue and Vanderburg
Volume: 527, Astronomical Data Analysis Software and Systems XXIX
Page: 187
Authors: Baldassi, L.; Brooks, A.
Abstract: This study partially replicated the results of the work of Shallue & Vanderburg (2018) who used deep learning for exoplanet identification on NASA Kepler Space Telescope light curve data. The software installation process is described including the versioning and assertion issues that occurred during this phase. Deep learning models were built under the convolutional neural network configuration and model performance was evaluated using the provided evaluation script. Results were similar to those previously published. A generated precision-recall curve is compared to the previously published curve. Around 1% of the light curves were unanalyzable for reasons that could not be established. As a check on software quality, static analysis of the source code was performed. The results showed many warnings could be classified as false positives. No serious code defect was revealed. Finally, the best performing model was applied to produce predictions over a set of 2421 unconfirmed Kepler Objects of Interest (KOI) from the Mikulski Archive for Space Telescopes labelled as planet candidates.
Back to Volume