Over 150 published articles across three decades, from astrophysics through data science to business analytics. Selected working papers and publications below.

Complete record: Google Scholar  ·  SSRN  ·  ORCID

Working Papers

Learning from Structure, Not Sequences: Forecasting Equity Markets Using GNNs on Financial Correlation Networks
Bracht, E., Brunner, R., & McMullin, J. L.
SSRN

Applies graph neural networks to financial correlation networks, demonstrating that network structure captures market dynamics more effectively than traditional time-series approaches for equity market forecasting.

Predicting Profitability Using Machine Learning
Anand, V., Brunner, R., Ikegwu, K., & Sougiannis, T. (2019).
SSRN

Demonstrates that machine learning methods substantially outperform traditional approaches in predicting future firm profitability, with ensemble techniques proving particularly effective.

Selected Publications

Towards a Unified Approach to Industry Recovery: Insights from Intraday Stock Data and Advanced Community Detection Methods
Bracht, E., McMullin, J. L., & Brunner, R. J. (2025).
Physica A: Statistical Mechanics and its Applications, 669, 130501.
Paper | SSRN

Explores how time series parameters—sampling frequency, sample period, and series length—affect recovery of industry classifications in financial networks. Using high-frequency S&P 500 data (2005-2012) with normalized mutual information and Planar Maximally Filtered Graphs, applies Leiden and spectral clustering to identify stock communities. Finds optimal industry structure recovery at 4-48 minute sampling frequencies, with clustering accuracy improving during volatile periods like the 2008 financial crisis.

Humans vs. ChatGPT: Evaluating Annotation Methods for Financial Corpora
Kaikaus, J., Li, H., & Brunner, R. J. (2023).
2023 IEEE International Conference on Big Data, pp. 2831-2838.
Paper | SSRN

Demonstrates that large language models (GPT-3.5 and GPT-4) provide more consistent and reliable annotations for emotion, sentiment, and cognitive dissonance in earnings calls than traditional human annotation methods, while being significantly more cost- and time-efficient.

Truth or Fiction: Multimodal Learning Applied to Earnings Calls
Kaikaus, J., Hobson, J. L., & Brunner, R. J. (2022).
2022 IEEE International Conference on Big Data.
Paper | SSRN

Develops a multimodal bidirectional LSTM framework with cross-attention fusion trained on audio and text from earnings calls, demonstrating that multimodal data substantially improves model accuracy for financial restatement prediction compared to text-only approaches.

Extended Isolation Forest
Hariri, S., Kind, M. C., & Brunner, R. J. (2021).
IEEE Transactions on Knowledge and Data Engineering, 33(4), 1479-1489.
Paper | arXiv:1811.02141GitHub repository

An extension to the model-free anomaly detection algorithm that resolves biases in anomaly score assignment using randomly oriented hyperplanes, dramatically improving consistency and reliability.

Star-Galaxy Classification Using Deep Convolutional Neural Networks
Kim, E. J., & Brunner, R. J. (2017).
Monthly Notices of the Royal Astronomical Society, 464(4), 4463-4475.
Paper | arXiv:1608.04369

Applied deep convolutional neural networks to astronomical image classification, demonstrating superior performance over traditional methods for distinguishing stars from galaxies in large-scale surveys.

Machine Learning and Cosmological Simulations I: Semi-Analytical Models
Kamdar, H. M., Turk, M. J., & Brunner, R. J. (2016).
Monthly Notices of the Royal Astronomical Society, 455(1), 642-658.
Paper | arXiv:1510.06402

Pioneering application of machine learning to model galaxy formation and evolution, demonstrating that ML algorithms can effectively predict galaxy properties using only dark matter halo characteristics and merger tree information.

Data Mining and Machine Learning in Astronomy
Ball, N. M., & Brunner, R. J. (2010).
International Journal of Modern Physics D, 19(7), 1049-1106.
Paper | arXiv:0906.2173

Comprehensive review of data mining and machine learning techniques in astronomy, covering algorithms, applications, and best practices for knowledge discovery in astronomical databases. One of the most cited review papers in the field.

Massive Datasets in Astronomy
Brunner, R. J., Djorgovski, S. G., Prince, T. A., & Szalay, A. S. (2001).
In Handbook of Massive Data Sets (pp. 931-979). Springer.
arXiv:astro-ph/0106481

Early vision for data mining and virtual observatories in astronomy, anticipating the data tsunami from modern surveys and establishing foundational approaches for handling massive astronomical datasets.