arXiv subjects

Brian Hutchinson

Publications and source records attributed to Brian Hutchinson.

24 records · Page 2Linked to original sources

arXiv2020

APOGEE Net: Improving the derived spectral parameters for young stars through deep learning

Machine learning allows efficient extraction of physical properties from stellar spectra that have been obtained by large surveys. The viability of ML approaches has been demonstrated for spectra covering a variety of wavelengths and spectral resolutions, but most often for main sequence or evolved stars, where reliable synthetic spectra provide labels and data for training. Spectral models of young stellar objects (YSOs) and low mass main sequence (MS) stars are less well-matched to their empirical counterparts, however, posing barriers to previous approaches to classify spectra of such stars. In this work we generate labels for YSOs and low mass MS stars through their photometry. We then use these labels to train a deep convolutional neural network to predict log g, Teff, and Fe/H for stars with APOGEE spectra in the DR14 dataset. This "APOGEE Net" has produced reliable predictions of log g for YSOs, with uncertainties of within 0.1 dex and a good agreement with the structure indicated by pre-main sequence evolutionary tracks, and correlate well with independently derived stellar radii. These values will be useful for studying pre-main sequence stellar populations to accurately diagnose membership and ages.

astro-ph.SR↗

arXiv2019

Metric-Based Few-Shot Learning for Video Action Recognition

In the few-shot scenario, a learner must effectively generalize to unseen classes given a small support set of labeled examples. While a relatively large amount of research has gone into few-shot learning for image classification, little work has been done on few-shot video classification. In this work, we address the task of few-shot video action recognition with a set of two-stream models. We evaluate the performance of a set of convolutional and recurrent neural network video encoder architectures used in conjunction with three popular metric-based few-shot algorithms. We train and evaluate using a few-shot split of the Kinetics 600 dataset. Our experiments confirm the importance of the two-stream setup, and find prototypical networks and pooled long short-term memory network embeddings to give the best performance as few-shot method and video encoder, respectively. For a 5-shot 5-way task, this setup obtains 84.2% accuracy on the test set and 59.4% on a special "challenge" test set, composed of highly confusable classes.

cs.CV↗

arXiv2018

Protein Mutation Stability Ternary Classification using Neural Networks and Rigidity Analysis

Discerning how a mutation affects the stability of a protein is central to the study of a wide range of diseases. Machine learning and statistical analysis techniques can inform how to allocate limited resources to the considerable time and cost associated with wet lab mutagenesis experiments. In this work we explore the effectiveness of using a neural network classifier to predict the change in the stability of a protein due to a mutation. Assessing the accuracy of our approach is dependent on the use of experimental data about the effects of mutations performed in vitro. Because the experimental data is prone to discrepancies when similar experiments have been performed by multiple laboratories, the use of the data near the juncture of stabilizing and destabilizing mutations is questionable. We address this later problem via a systematic approach in which we explore the use of a three-way classification scheme with stabilizing, destabilizing, and inconclusive labels. For a systematic search of potential classification cutoff values our classifier achieved 68 percent accuracy on ternary classification for cutoff values of -0.6 and 0.7 with a low rate of classifying stabilizing as destabilizing and vice versa.

q-bio.QM↗

arXiv2018

Recurrent Neural Network Attention Mechanisms for Interpretable System Log Anomaly Detection

Deep learning has recently demonstrated state-of-the art performance on key tasks related to the maintenance of computer systems, such as intrusion detection, denial of service attack detection, hardware and software system failures, and malware detection. In these contexts, model interpretability is vital for administrator and analyst to trust and act on the automated analysis of machine learning models. Deep learning methods have been criticized as black box oracles which allow limited insight into decision factors. In this work we seek to "bridge the gap" between the impressive performance of deep learning models and the need for interpretable model introspection. To this end we present recurrent neural network (RNN) language models augmented with attention for anomaly detection in system logs. Our methods are generally applicable to any computer system and logging source. By incorporating attention variants into our RNN language models we create opportunities for model introspection and analysis without sacrificing state-of-the art performance. We demonstrate model performance and illustrate model interpretability on an intrusion detection task using the Los Alamos National Laboratory (LANL) cyber security dataset, reporting upward of 0.99 area under the receiver operator characteristic curve despite being trained only on a single day's worth of data.

cs.LG↗

arXiv2017

Deep Learning for Unsupervised Insider Threat Detection in Structured Cybersecurity Data Streams

Analysis of an organization's computer network activity is a key component of early detection and mitigation of insider threat, a growing concern for many organizations. Raw system logs are a prototypical example of streaming data that can quickly scale beyond the cognitive power of a human analyst. As a prospective filter for the human analyst, we present an online unsupervised deep learning approach to detect anomalous network activity from system logs in real time. Our models decompose anomaly scores into the contributions of individual user behavior features for increased interpretability to aid analysts reviewing potential cases of insider threat. Using the CERT Insider Threat Dataset v6.2 and threat detection recall as our performance metric, our novel deep and recurrent neural network models outperform Principal Component Analysis, Support Vector Machine and Isolation Forest based anomaly detection baselines. For our best model, the events labeled as insider threat activity in our dataset had an average anomaly score in the 95.53 percentile, demonstrating our approach's potential to greatly reduce analyst workloads.

cs.NE↗

arXiv2017

Recurrent Neural Network Language Models for Open Vocabulary Event-Level Cyber Anomaly Detection

Automated analysis methods are crucial aids for monitoring and defending a network to protect the sensitive or confidential data it hosts. This work introduces a flexible, powerful, and unsupervised approach to detecting anomalous behavior in computer and network logs, one that largely eliminates domain-dependent feature engineering employed by existing methods. By treating system logs as threads of interleaved "sentences" (event log lines) to train online unsupervised neural network language models, our approach provides an adaptive model of normal network behavior. We compare the effectiveness of both standard and bidirectional recurrent neural network language models at detecting malicious activity within network log data. Extending these models, we introduce a tiered recurrent architecture, which provides context by modeling sequences of users' actions over time. Compared to Isolation Forest and Principal Components Analysis, two popular anomaly detection algorithms, we observe superior performance on the Los Alamos National Laboratory Cyber Security dataset. For log-line-level red team detection, our best performing character-based model provides test set area under the receiver operator characteristic curve of 0.98, demonstrating the strong fine-grained anomaly detection performance of this approach on open vocabulary logging sources.

cs.NE↗