Hossein Soleimani

Dr. Hossein Soleimani


Research Focus & Metrics

Academic Career Length

13Years (2014-Present)

(Calculated once Scilit publication data is synced: current year − year of first publication + 1)

Representative Publications

View All Publications
21 August 2017 in IEEE Transactions on Pattern Analysis and Machine Intelligence
Verified
Journal Article
Open access
HSHossein Soleimani
Hossein Soleimani
JHJames Hensman
James Hensman
SSSuchi Saria
Suchi Saria
Missing data and noisy observations pose significant challenges for reliably predicting events from irregularly sampled multivariate time series (longitudinal) data. Imputation methods, which are typically used for completing the data prior to event prediction, lack a principled mechanism to account for the uncertainty due to missingness. Alternatively, state-of-the-art joint modeling techniques can be used for jointly modeling the longitudinal and event data and compute event probabilities conditioned on the longitudinal observations. These approaches, however, make strong parametric assumptions and do not easily scale to multivariate signals with many observations. Our proposed approach consists of several key innovations. First, we develop a flexible and scalable joint model based upon sparse multiple-output Gaussian processes. Unlike state-of-the-art joint models, the proposed model can explain highly challenging structure including non-Gaussian noise while scaling to large data. Second, we derive an optimal policy for predicting events using the distribution of the event occurrence estimated by the joint model. The derived policy trades-off the cost of a delayed detection versus incorrect assessments and abstains from making decisions when the estimated event probability does not satisfy the derived confidence criteria. Experiments on a large dataset show that the proposed framework significantly outperforms state-of-the-art techniques in event prediction.

1 April 2017 in Neural Computation
Verified
Journal Article
Closed access
HSHossein Soleimani
Hossein Soleimani
DMDavid J. Miller
David J. Miller
Many classification tasks require both labeling objects and determining label associations for parts of each object. Example applications include labeling segments of images or determining relevant parts of a text document when the training labels are available only at the image or document level. This task is usually referred to as multi-instance (MI) learning, where the learner typically receives a collection of labeled (or sometimes unlabeled) bags, each containing several segments (instances). We propose a semisupervised MI learning method for multilabel classification. Most MI learning methods treat instances in each bag as independent and identically distributed samples. However, in many practical applications, instances are related to each other and should not be considered independent. Our model discovers a latent low-dimensional space that captures structure within each bag. Further, unlike many other MI learning methods, which are primarily developed for binary classification, we model multiple ...


26 November 2025 in NEJM AI
Verified
Journal Article
Closed access
ABAndrew Bishara
Andrew Bishara
Marlene LinMarlene Lin
Marlene Lin
JDJean Digitale
Jean Digitale
JYJoanne W. L. Yim
Joanne W. L. Yim
HSHossein Soleimani
Hossein Soleimani
PRPriya Ramaswamy
Priya Ramaswamy
Current preoperative blood ordering relies on outdated, procedure-based maximum surgical blood order schedule (MSBOS) that overlook individual patient risk. Smart Match, a machine-learning (ML) tool that predicts patient-specific transfusion needs, was developed and silently validated. Integrated into real-time workflows, Smart Match demonstrates the potential of ML to replace MSBOS with personalized, data-driven approaches to perioperative blood management. Using 82 variables and 1921 features from the electronic medical record of a tertiary academic medical center, a custom extreme gradient boosting (XGBoost) model was optimized to predict perioperative red blood cell (RBC) transfusions before performing adult elective surgeries, aligning sensitivity with native MSBOS thresholds. The retrospective cohort was split into training, validation, and test sets. Silent prospective validation was conducted in real-world elective cases, comparing Smart Match predictions with both MSBOS and clinician RBC preordering behaviors at multiple time points preceding surgery. Model discrimination, calibration, and clinical utility were assessed. In addition, a hybrid model to estimate daily blood bank needs was developed. Retrospective data consisted of 235,054 cases with a 3.04% (95% confidence interval [CI], 2.97 to 3.11) transfusion rate. The model predictors included medical history, laboratory results, demographics, surgery information, medications, transfusion history, and MSBOS recommendations. The test data achieved an area under the receiver-operating-characteristic curve (AUROC) of 0.94 (95% CI, 0.93 to 0.95) and an area under the precision–recall curve (AUPRC) of 0.57 (95% CI, 0.53 to 0.61). Silent prospective validation (n=24,003, transfusion rate 2.18%, 95% CI, 2.07 to 2.46) maintained a 0.94 (95% CI, 0.92 to 0.95) AUROC and a 0.55 (95% CI, 0.51 to 0.60) AUPRC. Prospectively, the model’s sensitivity was 0.72 (95% CI, 0.67 to 0.75) and its positive predictive value was 0.34 (95% CI, 0.31 to 0.37), surpassing both MSBOS and clinician behavior. Our hybrid model outperformed MSBOS for daily RBC needs with a mean absolute error of 12.86 (95% CI, 11.86 to 13.84) versus 13.34 (95% CI, 12.14 to 14.51) on the test set. Smart Match reliably outperforms MSBOS and clinician orders, works in real time, and will next be assessed through a randomized control trial to gauge its usability and clinical impact. (Funded by the National Institute of General Medical Sciences and others.)

11 September 2014 in arXiv
Verified
Preprint
Open access
PDF file
HSHossein Soleimani
Hossein Soleimani
DMDavid Miller
David Miller
We propose a parsimonious topic model for text corpora. In related models such as Latent Dirichlet Allocation (LDA), all words are modeled topic-specifically, even though many words occur with similar frequencies across different topics. Our modeling determines salient words for each topic, which have topic-specific probabilities, with the rest explained by a universal shared model. Further, in LDA all topics are in principle present in every document. By contrast our model gives sparse topic representation, determining the (small) subset of relevant topics for each document. We derive a Bayesian Information Criterion (BIC), balancing model complexity and goodness of fit. Here, interestingly, we identify an effective sample size and corresponding penalty specific to each parameter type in our model. We minimize BIC to jointly determine our entire model -- the topic-specific words, document-specific topics, all model parameter values, {\it and} the total number of topics -- in a wholly unsupervised fashion. Results on three text corpora and an image dataset show that our model achieves higher test set likelihood and better agreement with ground-truth class labels, compared to LDA and to a model designed to incorporate sparsity.

20 May 2016 in arXiv
Verified
Preprint
Open access
PDF file
HSHossein Soleimani
Hossein Soleimani
DMDavid Miller
David Miller
We propose an algorithm for detecting patterns exhibited by anomalous clusters in high dimensional discrete data. Unlike most anomaly detection (AD) methods, which detect individual anomalies, our proposed method detects groups (clusters) of anomalies; i.e. sets of points which collectively exhibit abnormal patterns. In many applications this can lead to better understanding of the nature of the atypical behavior and to identifying the sources of the anomalies. Moreover, we consider the case where the atypical patterns exhibit on only a small (salient) subset of the very high dimensional feature space. Individual AD techniques and techniques that detect anomalies using all the features typically fail to detect such anomalies, but our method can detect such instances collectively, discover the shared anomalous patterns exhibited by them, and identify the subsets of salient features. In this paper, we focus on detecting anomalous topics in a batch of text documents, developing our algorithm based on topic models. Results of our experiments show that our method can accurately detect anomalous topics and salient features (words) under each such topic in a synthetic data set and two real-world text corpora and achieves better performance compared to both standard group AD and individual AD techniques. All required code to reproduce our experiments is available from https://github.com/hsoleimani/ATD