Sahib Julka

The magnetosphere of Mercury is relatively small and highly dynamic, mostly due to the weak planetary magnetic field. Varying solar wind conditions principally determine the location of both the Hermean bow shock and magnetopause. In 2011 – 2015 MESSENGER spacecraft completed over 4000 orbits around Mercury, thus giving a data of more than 8000 crossings of bow shock and magnetopause of the planet, this makes it possible to study in detail the bow shock, the magnetopause and the magnetosheath structures.In this work we determine crossings of the bow shock and the magnetopause of Mercury by applying machine learning methods to the MESSENGER magnetometer data. We attempt to identify the crossings during the whole duration of the orbital mission and model the average three-dimensional shapes of these boundaries. The results are compared with those previously obtained in other works.This work may be of interest for future Mercury research related to the BepiColombo spacecraft mission, which will enter the orbit around the planet in December 2025.

In real-world applications, input data distributions are rarely static over a period of time, a phenomenon known as concept drift. Such concept drifts degrade the model's prediction performance, and therefore we require methods to overcome these issues. The initial step is to identify concept drifts and have a training method in place to recover the model's performance. Most concept drift detection methods work on detecting concept drifts and signalling the requirement to retrain the model. However, in real-world cases, there could be concept drifts that recur over a period of time. In this paper, we present an unsupervised method based on Generative Adversarial Networks(GAN) to detect concept drifts and identify whether a specific concept drift occurred in the past. Our method reduces the time and data the model requires to get up to speed for recurring drifts. Our key results indicate that our proposed model can outperform the current state-of-the-art models in most datasets. We also test our method on a real-world use case from astrophysics, where we detect the bow shock and magnetopause crossings with better results than the existing methods in the domain.

The last decade has witnessed a rapid growth of the field of exoplanet discovery and characterization. However, several big challenges remain, many of which could be addressed using machine learning methodology. For instance, the most prolific method for detecting exoplanets and inferring several of their characteristics, transit photometry, is very sensitive to the presence of stellar spots. The current practice in the literature is identifying the effects of spots visually and correcting them manually or discarding the affected data. This paper explores a first step towards fully automating the efficient and precise derivation of transit depths from transit light curves in the presence of stellar spots. The primary focus of the paper is to present in detail a diverse arsenal of methods for doing so. The methods and results we present were obtained in the context of the 1st Machine Learning Challenge organized for the European Space Agency’s upcoming Ariel mission. We first present the problem, the simulated Ariel-like data and outline the Challenge while identifying best practices for organizing similar challenges in the future. Finally, we present the solutions obtained by the top five winning teams, provide their code, and discuss their implications. Successful solutions either construct highly non-linear (w.r.t. the raw data) models with minimal pre-processing – deep neural networks and ensemble methods – or amount to obtaining meaningful statistics from the light curves, constructing linear models on which yields comparably good predictive performance.

The goal of this work is to improve our understanding of Mercury's magnetosphere and its dynamics. The magnetosphere of Mercury is rather small and highly dynamic due to its weak internal magnetic field and its close proximity to the Sun. The changing solar wind conditions principally determine the locations of both the Hermean bow shock and magnetopause. In 2011 – 2015 MESSENGER (MErcury Surface, Space ENvironment, GEochemistry, and Ranging) spacecraft completed over 4000 orbits around Mercury, producing a large dataset of magnetic field measurements, which have subsequently been manually annotated in the present work as well as in other works such as Philpott et al. [1] to determine magnetopause and bow shock crossings. The size of the dataset enables usage of supervised machine learning algorithms to detect such crossings in archive data and predict them in real time.Pre-processed features are first windowed using a sliding window of variable size, normalized and into a block of 3 Convolutional layers with 1D filters, each followed by Batch Normalization and ReLu activations. The activations obtained at the end of the CNN block are then passed to the Recurrent block with two layers of LSTMs. The final activations are then passed to a fully connected layer with softmax activations. The objective function used for training is Categorical cross entropy, with Adam optimizer. Overall, the predictor achieves a macro F1 score of about 80% on the Bow-shock and the magnetopause crossings on a randomly sampled test of 300 MESSENGER orbits (two mercury years of data).This work may be of particular interest for future Mercury research related to the BepiColombo spacecraft mission, which will enter orbit around the planet in December 2025.[1] L. C. Philpott et al. “The Shape of Mercury’s Magnetopause: The Picture From MESSENGER Magnetometer Observations and Future Prospects for BepiColombo”. In: Journal of Geophysical Research: Space Physics 125.5 (May 2020). issn: 2169-9380, 2169-9402. doi: 10.1029/2019JA027544.