Arrow Research search

Author name cluster

Carla E. Brodley

Possible papers associated with this exact author name in Arrow. This page groups case-insensitive exact name matches and is not a full identity disambiguation profile.

15 papers
2 author rows

Possible papers

15

ICLR Conference 2025 Conference Paper

NNsight and NDIF: Democratizing Access to Open-Weight Foundation Model Internals

  • Jaden Fried Fiotto-Kaufman
  • Alexander Russell Loftus
  • Eric Todd
  • Jannik Brinkmann
  • Koyena Pal
  • Dmitrii Troitskii
  • Michael Ripa
  • Adam Belfki

We introduce NNsight and NDIF, technologies that work in tandem to enable scientific study of the representations and computations learned by very large neural networks. NNsight is an open-source system that extends PyTorch to introduce deferred remote execution. The National Deep Inference Fabric (NDIF) is a scalable inference service that executes NNsight requests, allowing users to share GPU resources and pretrained models. These technologies are enabled by the Intervention Graph, an architecture developed to decouple experimental design from model runtime. Together, this framework provides transparent and efficient access to the internals of deep neural networks such as very large language models (LLMs) without imposing the cost or complexity of hosting customized models individually. We conduct a quantitative survey of the machine learning literature that reveals a growing gap in the study of the internals of large-scale AI. We demonstrate the design and use of our framework to address this gap by enabling a range of research methods on huge models. Finally, we conduct benchmarks to compare performance with previous approaches. Code, documentation, and tutorials are available at https://nnsight.net/.

Details

JMLR Journal 2016 Journal Article

Decrypting “Cryptogenic” Epilepsy: Semi-supervised Hierarchical Conditional Random Fields For Detecting Cortical Lesions In MRI-Negative Patients

  • Bilal Ahmed
  • Thomas Thesen
  • Karen E. Blackmon
  • Ruben Kuzniekcy
  • Orrin Devinsky
  • Carla E. Brodley

Focal cortical dysplasia (FCD) is the most common cause of pediatric epilepsy and the third most common cause in adults with treatment-resistant epilepsy. Surgical resection of the lesion is the most effective treatment to stop seizures. Technical advances in MRI have revolutionized the diagnosis of FCD, leading to high success rates for resective surgery. However, 45% of histologically confirmed FCD patients have normal MRIs ( MRI-negative ). Without a visible lesion, the success rate of surgery drops from 66% to 29%. In this work, we cast the problem of detecting potential FCD lesions using MRI scans of MRI-negative patients in an image segmentation framework based on hierarchical conditional random fields (HCRF). We use surface based morphometry to model the cortical surface as a two-dimensional surface which is then segmented at multiple scales to extract superpixels of different sizes. Each superpixel is assigned an outlier score by comparing it to a control population. The lesion is detected by fusing the outlier probabilities across multiple scales using a tree- structured HCRF. The proposed method achieves a higher detection rate, with superior recall and precision on a sample of twenty MRI-negative FCD patients as compared to a baseline across four morphological features and their combinations. [abs] [ pdf ][ bib ] &copy JMLR 2016. ( edit, beta )

PDF Details

AIIM Journal 2015 Journal Article

Removing confounding factors via constraint-based clustering: An application to finding homogeneous groups of multiple sclerosis patients

  • Jingjing Liu
  • Carla E. Brodley
  • Brian C. Healy
  • Tanuja Chitnis

Objectives Confounding factors in unsupervised data can lead to undesirable clustering results. For example in medical datasets, age is often a confounding factor in tests designed to judge the severity of a patient's disease through measures of mobility, eyesight and hearing. In such cases, removing age from each instance will not remove its effect from the data as other features will be correlated with age. Motivated by the need to find homogeneous groups of multiple sclerosis (MS) patients, we apply our approach to remove physician subjectivity from patient data. Methods We present a method based on constraint-based clustering to remove the impact of such confounding factors. Given knowledge about which feature (or set of features) is a confounding factor, call it F. Our method first partitions the data into b bins: if F is categorical, instances from the same category construct one bin; if F is numeric, then we split bins such that each bin contains instances of similar F value. Thus each instance is assigned to a single bin for factor F. We then remove feature F from each instance for the remaining steps. Next, we cluster the data separately in each bin. Using these clustering results, we generate pair-wise constraints and then run a constraint-based clustering algorithm to produce a final grouping. Results In a series of experiments with synthetic datasets, we compare our proposed methods to detrending when one has numeric confounding factors. We apply our method to the Comprehensive Longitudinal Investigation of Multiple Sclerosis at Brigham and Womens Hospital dataset, and find a novel grouping of patients that can help uncover the factors that impact disease progression in MS. Conclusions Our method groups data removing the effect of confounding factors without making any assumptions about the form of the influence of these factors on the other features. We identified clusters of MS patients that have clinically recognizable differences. Because patients more likely to progress are found using this approach, our results have the potential to aid physicians in tailoring treatment decisions for MS patients.

Details DOI

ICML Conference 2014 Conference Paper

Hierarchical Conditional Random Fields for Outlier Detection: An Application to Detecting Epileptogenic Cortical Malformations

  • Bilal Ahmed
  • Thomas Thesen
  • Karen E. Blackmon
  • Yijun Zhao
  • Orrin Devinsky
  • Ruben Kuzniecky
  • Carla E. Brodley

We cast the problem of detecting and isolating regions of abnormal cortical tissue in the MRIs of epilepsy patients in an image segmentation framework. Employing a multiscale approach we divide the surface images into segments of different sizes and then classify each segment as being an outlier, by comparing it to the same region across controls. The final classification is obtained by fusing the outlier probabilities obtained at multiple scales using a tree-structured hierarchical conditional random field (HCRF). The proposed method correctly detects abnormal regions in 90% of patients whose abnormality was detected via routine visual inspection of their clinical MRI. More importantly, it detects abnormalities in 80% of patients whose abnormality escaped visual inspection by expert radiologists.

Details

JMLR Journal 2004 Journal Article

Feature Selection for Unsupervised Learning

  • Jennifer G. Dy
  • Carla E. Brodley

In this paper, we identify two issues involved in developing an automated feature subset selection algorithm for unlabeled data: the need for finding the number of clusters in conjunction with feature selection, and the need for normalizing the bias of fe ature selection criteria with respect to dimension. We explore the feature selection problem and these issues through FSSEM (Feature Subset Selection using Expectation-Maximization (EM) clustering) and through two different performance criteria for evaluating candidate feature subsets: scatter separability and maximum likelihood. We present proofs on the dimensionality biases of these feature criteria, and present a cross-projection normalization scheme that can be applied to any criterion to ameliorate these biases. Our experiments show the need for feature selection, the need for addressing these two issues, and the effectiveness of our proposed solutions. [abs] [ pdf ][ ps.gz ][ ps ]

PDF Details

ICML Conference 2004 Conference Paper

Solving cluster ensemble problems by bipartite graph partitioning

  • Xiaoli Z. Fern
  • Carla E. Brodley

A critical problem in cluster ensemble research is how to combine multiple clusterings to yield a final superior clustering result. Leveraging advanced graph partitioning techniques, we solve this problem by reducing it to a graph partitioning problem. We introduce a new reduction method that constructs a bipartite graph from a given cluster ensemble. The resulting graph models both instances and clusters of the ensemble simultaneously as vertices in the graph. Our approach retains all of the information provided by a given ensemble, allowing the similarity among instances and the similarity among clusters to be considered collectively in forming the final clustering. Further, the resulting graph partitioning problem can be solved efficiently. We empirically evaluate the proposed approach against two commonly used graph formulations and show that it is more robust and achieves comparable or better performance in comparison to its competitors.

Details

AAAI Conference 1996 Conference Paper

Identifying and Eliminating Mislabeled Training Instances

  • Carla E. Brodley

This paper presents a new approach to identifying and eliminating mislabeled training instances. The goal of this technique is to improve classification accuracies produced by learning algorithms by improving the quality of the training data. The approach employs an ensemble of classifiers that serve as a filter for the training data. Using an n-fold cross validation, the training data is passed through the filter. Only instances that the filter classifies correctly are passed to the final learning algorithm. We present an empirical evaluation of the approach for the task of automated land cover mapping from remotely sensed data. Labeling error arises in these data from a multitude of sources including lack of consistency in the vegetation classification used, variable measurement techniques, and variation in the spatial sampling resolution. Our evaluation shows that for noise levels of less than 40%, filtering results in higher predictive accuracy than not filtering, and for levels of class noise less than or equal to 20% filtering allows the base-line accuracy to be retained. Our empirical results suggest that the ensemble filter approach is an effective method for identifying labeling errors, and further, that the approach will significantly benefit ongoing research to develop accurate and robust remote sensing-based methods to map land cover at global scales.

PDF Details
v2026.09.13