Sanjay Purushotham

Catherine Ordun, Alexandra N. Cha, Edward Raff et al.

Cancer patients experience high rates of chronic pain throughout the treatment process. Assessing pain for this patient population is a vital component of psychological and functional well-being, as it can cause a rapid deterioration of quality of life. Existing work in facial pain detection often have deficiencies in labeling or methodology that prevent them from being clinically relevant. This paper introduces the first chronic cancer pain dataset, collected as part of the Intelligent Sight and Sound (ISS) clinical trial, guided by clinicians to help ensure that model findings yield clinically relevant results. The data collected to date consists of 29 patients, 509 smartphone videos, 189,999 frames, and self-reported affective and activity pain scores adopted from the Brief Pain Inventory (BPI). Using static images and multi-modal data to predict self-reported pain levels, early models show significant gaps between current methods available to predict pain today, with room for improvement. Due to the especially sensitive nature of the inherent Personally Identifiable Information (PII) of facial images, the dataset will be released under the guidance and control of the National Institutes of Health (NIH).

Seraj Al Mahmud Mostafa, Omar Faruque, Chenxi Wang et al.

Atmospheric gravity waves occur in the Earths atmosphere caused by an interplay between gravity and buoyancy forces. These waves have profound impacts on various aspects of the atmosphere, including the patterns of precipitation, cloud formation, ozone distribution, aerosols, and pollutant dispersion. Therefore, understanding gravity waves is essential to comprehend and monitor changes in a wide range of atmospheric behaviors. Limited studies have been conducted to identify gravity waves from satellite data using machine learning techniques. Particularly, without applying noise removal techniques, it remains an underexplored area of research. This study presents a novel kernel design aimed at identifying gravity waves within satellite images. The proposed kernel is seamlessly integrated into a deep convolutional neural network, denoted as gWaveNet. Our proposed model exhibits impressive proficiency in detecting images containing gravity waves from noisy satellite data without any feature engineering. The empirical results show our model outperforms related approaches by achieving over 98% training accuracy and over 94% test accuracy which is known to be the best result for gravity waves detection up to the time of this work. We open sourced our code at https://rb.gy/qn68ku.

Catherine Ordun, Alexandra Cha, Edward Raff et al.

Since thermal imagery offers a unique modality to investigate pain, the U.S. National Institutes of Health (NIH) has collected a large and diverse set of cancer patient facial thermograms for AI-based pain research. However, differing angles from camera capture between thermal and visible sensors has led to misalignment between Visible-Thermal (VT) images. We modernize the classic computer vision task of image registration by applying and modifying a generative alignment algorithm to register VT cancer faces, without the need for a reference or alignment parameters. By registering VT faces, we demonstrate that the quality of thermal images produced in the generative AI downstream task of Visible-to-Thermal (V2T) image translation significantly improves up to 52.5\%, than without registration. Images in this paper have been approved by the NIH NCI for public dissemination.

Catherine Ordun, Edward Raff, Sanjay Purushotham

Thermal facial imagery offers valuable insight into physiological states such as inflammation and stress by detecting emitted radiation in the infrared spectrum, which is unseen in the visible spectra. Telemedicine applications could benefit from thermal imagery, but conventional computers are reliant on RGB cameras and lack thermal sensors. As a result, we propose the Visible-to-Thermal Facial GAN (VTF-GAN) that is specifically designed to generate high-resolution thermal faces by learning both the spatial and frequency domains of facial regions, across spectra. We compare VTF-GAN against several popular GAN baselines and the first conditional Denoising Diffusion Probabilistic Model (DDPM) for VT face translation (VTF-Diff). Results show that VTF-GAN achieves high quality, crisp, and perceptually realistic thermal faces using a combined set of patch, temperature, perceptual, and Fourier Transform losses, compared to all baselines including diffusion.

Zahid Hasan, Masud Ahmed, Abu Zaher Md Faridee et al.

Novel Categories Discovery (NCD) facilitates learning from a partially annotated label space and enables deep learning (DL) models to operate in an open-world setting by identifying and differentiating instances of novel classes based on the labeled data notions. One of the primary assumptions of NCD is that the novel label space is perfectly disjoint and can be equipartitioned, but it is rarely realized by most NCD approaches in practice. To better align with this assumption, we propose a novel single-stage joint optimization-based NCD method, Negative learning, Entropy, and Variance regularization NCD (NEV-NCD). We demonstrate the efficacy of NEV-NCD in previously unexplored NCD applications of video action recognition (VAR) with the public UCF101 dataset and a curated in-house partial action-space annotated multi-view video dataset. We perform a thorough ablation study by varying the composition of final joint loss and associated hyper-parameters. During our experiments with UCF101 and multi-view action dataset, NEV-NCD achieves ~ 83% classification accuracy in test instances of labeled data. NEV-NCD achieves ~ 70% clustering accuracy over unlabeled data outperforming both naive baselines (by ~ 40%) and state-of-the-art pseudo-labeling-based approaches (by ~ 3.5%) over both datasets. Further, we propose to incorporate optional view-invariant feature learning with the multiview dataset to identify novel categories from novel viewpoints. Our additional view-invariance constraint improves the discriminative accuracy for both known and unknown categories by ~ 10% for novel viewpoints.

Md Mahmudur Rahman, Sanjay Purushotham

Survival analysis, time-to-event analysis, is an important problem in healthcare since it has a wide-ranging impact on patients and palliative care. Many survival analysis methods have assumed that the survival data is centrally available either from one medical center or by data sharing from multi-centers. However, the sensitivity of the patient attributes and the strict privacy laws have increasingly forbidden sharing of healthcare data. To address this challenge, the research community has looked at the solution of decentralized training and sharing of model parameters using the Federated Learning (FL) paradigm. In this paper, we study the utilization of FL for performing survival analysis on distributed healthcare datasets. Recently, the popular Cox proportional hazard (CPH) models have been adapted for FL settings; however, due to its linearity and proportional hazards assumptions, CPH models result in suboptimal performance, especially for non-linear, non-iid, and heavily censored survival datasets. To overcome the challenges of existing federated survival analysis methods, we leverage the predictive accuracy of the deep learning models and the power of pseudo values to propose a first-of-its-kind, pseudo value-based deep learning model for federated survival analysis (FSA) called FedPseudo. Furthermore, we introduce a novel approach of deriving pseudo values for survival probability in the FL settings that speeds up the computation of pseudo values. Extensive experiments on synthetic and real-world datasets show that our pseudo valued-based FL framework achieves similar performance as the best centrally trained deep survival analysis model. Moreover, our proposed FL approach obtains the best results for various censoring settings.

Catherine Ordun, Edward Raff, Sanjay Purushotham

For a variety of biometric cross-spectral tasks, Visible-Thermal (VT) facial pairs are used. However, due to a lack of calibration in the lab, photographic capture between two different sensors leads to severely misaligned pairs that can lead to poor results for person re-identification and generative AI. To solve this problem, we introduce our approach for VT image registration called Vista Morph. Unlike existing VT facial registration that requires manual, hand-crafted features for pixel matching and/or a supervised thermal reference, Vista Morph is completely unsupervised without the need for a reference. By learning the affine matrix through a Vision Transformer (ViT)-based Spatial Transformer Network (STN) and Generative Adversarial Networks (GAN), Vista Morph successfully aligns facial and non-facial VT images. Our approach learns warps in Hard, No, and Low-light visual settings and is robust to geometric perturbations and erasure at test time. We conduct a downstream generative AI task to show that registering training data with Vista Morph improves subject identity of generated thermal faces when performing V2T image translation.

Zahid Hasan, Abu Zaher Md Faridee, Masud Ahmed et al.

Novel Categories Discovery (NCD) aims to cluster novel data based on the class semantics of known classes using the open-world partial class space annotated dataset. As an alternative to the traditional pseudo-labeling-based approaches, we leverage the connection between the data sampling and the provided multinoulli (categorical) distribution of novel classes. We introduce constraints on individual and collective statistics of predicted novel class probabilities to implicitly achieve semantic-based clustering. More specifically, we align the class neuron activation distributions under Monte-Carlo sampling of novel classes in large batches by matching their empirical first-order (mean) and second-order (covariance) statistics with the multinoulli distribution of the labels while applying instance information constraints and prediction consistency under label-preserving augmentations. We then explore a directional statistics-based probability formation that learns the mixture of Von Mises-Fisher distribution of class labels in a unit hypersphere. We demonstrate the discriminative ability of our approach to realize semantic clustering of novel samples in image, video, and time-series modalities. We perform extensive ablation studies regarding data, networks, and framework components to provide better insights. Our approach maintains 94%, 93%, 85%, and 93% (approx.) classification accuracy in labeled data while achieving 90%, 84%, 72%, and 75% (approx.) clustering accuracy for novel categories in Cifar10, UCF101, MPSC-ARL, and SHAR datasets that match state-of-the-art approaches without any external clustering.

Md Mahmudur Rahman, Sanjay Purushotham

Multi-state survival analysis (MSA) uses multi-state models for the analysis of time-to-event data. In medical applications, MSA can provide insights about the complex disease progression in patients. A key challenge in MSA is the accurate subject-specific prediction of multi-state model quantities such as transition probability and state occupation probability in the presence of censoring. Traditional multi-state methods such as Aalen-Johansen (AJ) estimators and Cox-based methods are respectively limited by Markov and proportional hazards assumptions and are infeasible for making subject-specific predictions. Neural ordinary differential equations for MSA relax these assumptions but are computationally expensive and do not directly model the transition probabilities. To address these limitations, we propose a new class of pseudo-value-based deep learning models for multi-state survival analysis, where we show that pseudo values - designed to handle censoring - can be a natural replacement for estimating the multi-state model quantities when derived from a consistent estimator. In particular, we provide an algorithm to derive pseudo values from consistent estimators to directly predict the multi-state survival quantities from the subject's covariates. Empirical results on synthetic and real-world datasets show that our proposed models achieve state-of-the-art results under various censoring settings.

Bahirah Adewunmi, Edward Raff, Sanjay Purushotham

Automating network security analysis, particularly the identification of potential attack paths, presents significant challenges. Due in part to the sequential, interconnected, and evolutionary nature of system events which most artificial intelligence (AI) techniques struggle to model effectively. This paper proposes a Reinforcement Learning (RL) environment generation framework that simulates the sequence of processes executed on a Windows operating system, enabling dynamic modeling of malicious processes on a system. This methodology models operating system state and transitions using a graph representation. This graph is derived from open-source System Monitor (Sysmon) logs. To address the variety in system event types, fields, and log formats, a mechanism was developed to capture and model parent-child processes from Sysmon logs. A Gymnasium environment (SubstratumGraphEnv) was constructed to establish the perceptible basis for an RL environment, and a customized PyTorch interface was also built (SubstratumBridge) to translate Gymnasium graphs into Deep Reinforcement Learning (DRL) observations and discrete actions. Graph Convolutional Networks (GCNs) concretize the graph's local and global state, which feed the distinct policy and critic heads of an Advantage Actor-Critic (A2C) model. This work's central contribution lies in the design of a novel deep graphical RL environment that automates translation of sequential user and system events, furnishing crucial context for cybersecurity analysis. This work provides a foundation for future research into shaping training parameters and advanced reward shaping, while also offering insight into which system events attributes are critical to training autonomous RL agents.

Zahid Hassan Tushar, Sanjay Purushotham

The NASA PACE mission provides unprecedented hyperspectral observations of ocean color, aerosols, and clouds, offering new insights into how these components interact and influence Earth's climate and air quality. Its Ocean Color Instrument measures light across hundreds of finely spaced wavelength bands, enabling detailed characterization of features such as phytoplankton composition, aerosol properties, and cloud microphysics. However, hyperspectral data of this scale is large, complex, and difficult to label, requiring specialized processing and analysis techniques. Existing foundation models, which have transformed computer vision and natural language processing, are generally trained on standard RGB imagery and therefore struggle to interpret the continuous spectral signatures captured by PACE. While recent advances have introduced hyperspectral foundation models, they are typically trained on cloud-free observations and often remain limited to single-sensor datasets due to spectral inconsistencies across instruments. Moreover, existing models tend to be parameter-heavy and computationally expensive, limiting scalability and adoption in operational settings. To address these challenges, we introduce HyperFM, a parameter-efficient hyperspectral foundation model that leverages intra-group and inter-group spectral attention along with hybrid parameter decomposition to better capture spectral spatial relationships while reducing computational cost. HyperFM demonstrates consistent performance improvements over existing hyperspectral foundation models and task-specific state-of-the-art methods across four benchmark downstream atmospheric cloud property retrieval tasks. To support further research, we additionally release HyperFM250K, a large-scale hyperspectral dataset from the PACE mission that includes both clear and cloudy scenes.

Masud Ahmed, Zahid Hasan, Syed Arefinul Haque et al.

Traditional transformer-based semantic segmentation relies on quantized embeddings. However, our analysis reveals that autoencoder accuracy on segmentation mask using quantized embeddings (e.g. VQ-VAE) is 8% lower than continuous-valued embeddings (e.g. KL-VAE). Motivated by this, we propose a continuous-valued embedding framework for semantic segmentation. By reformulating semantic mask generation as a continuous image-to-embedding diffusion process, our approach eliminates the need for discrete latent representations while preserving fine-grained spatial and semantic details. Our key contribution includes a diffusion-guided autoregressive transformer that learns a continuous semantic embedding space by modeling long-range dependencies in image features. Our framework contains a unified architecture combining a VAE encoder for continuous feature extraction, a diffusion-guided transformer for conditioned embedding generation, and a VAE decoder for semantic mask reconstruction. Our setting facilitates zero-shot domain adaptation capabilities enabled by the continuity of the embedding space. Experiments across diverse datasets (e.g., Cityscapes and domain-shifted variants) demonstrate state-of-the-art robustness to distribution shifts, including adverse weather (e.g., fog, snow) and viewpoint variations. Our model also exhibits strong noise resilience, achieving robust performance ($\approx$ 95% AP compared to baseline) under gaussian noise, moderate motion blur, and moderate brightness/contrast variations, while experiencing only a moderate impact ($\approx$ 90% AP compared to baseline) from 50% salt and pepper noise, saturation and hue shifts. Code available: https://github.com/mahmed10/CAMSS.git

Alabi Ahmed, Vandana Janeja, Sanjay Purushotham

The rapid advances in text-to-speech (TTS) technologies have made audio deepfakes increasingly realistic and accessible, raising significant security and trust concerns. While existing research has largely focused on detecting single-speaker audio deepfakes, real-world malicious applications with multi-speaker conversational settings is also emerging as a major underexplored threat. To address this gap, we propose a conceptual taxonomy of multi-speaker conversational audio deepfakes, distinguishing between partial manipulations (one or multiple speakers altered) and full manipulations (entire conversations synthesized). As a first step, we introduce a new Multi-speaker Conversational Audio Deepfakes Dataset (MsCADD) of 2,830 audio clips containing real and fully synthetic two-speaker conversations, generated using VITS and SoundStorm-based NotebookLM models to simulate natural dialogue with variations in speaker gender, and conversational spontaneity. MsCADD is limited to text-to-speech (TTS) types of deepfake. We benchmark three neural baseline models; LFCC-LCNN, RawNet2, and Wav2Vec 2.0 on this dataset and report performance in terms of F1 score, accuracy, true positive rate (TPR), and true negative rate (TNR). Results show that these baseline models provided a useful benchmark, however, the results also highlight that there is a significant gap in multi-speaker deepfake research in reliably detecting synthetic voices under varied conversational dynamics. Our dataset and benchmarks provide a foundation for future research on deepfake detection in conversational scenarios, which is a highly underexplored area of research but also a major area of threat to trustworthy information in audio settings. The MsCADD dataset is publicly available to support reproducibility and benchmarking by the research community.

Zahid Hassan Tushar, Sanjay Purushotham

Aerosol Optical Depth (AOD) retrieval is essential for Earth observation, supporting applications from air quality monitoring to climate studies. Conventional physics-based AOD retrieval methods formulate the problem as a pixel-wise inversion, relying on radiative transfer modeling, memory-intensive look-up tables, and auxiliary meteorological data. While recent data-driven approaches have shown promise, many fail to exploit the spatial-spectral coherence of hyperspectral imagery, leading to spatially inconsistent and noise-sensitive retrievals. We present the first study exploring Foundation AI models for AOD retrieval and propose ViTCG, a Vision Transformer with Channel-wise Grouping-based spatial regression framework that reduces retrieval bias and error. ViTCG uses hyperspectral top-of-atmosphere radiance as input and jointly models spatial context and spectral information. Validation with PACE radiance observations demonstrates a 62% reduction in mean squared error compared to state-of-the-art foundation models, including Prithvi, and produces spatially coherent AOD fields.

Elizabeth G. Campolongo, Yuan-Tang Chou, Ekaterina Govorkova et al.

Scientific discoveries are often made by finding a pattern or object that was not predicted by the known rules of science. Oftentimes, these anomalous events or objects that do not conform to the norms are an indication that the rules of science governing the data are incomplete, and something new needs to be present to explain these unexpected outliers. The challenge of finding anomalies can be confounding since it requires codifying a complete knowledge of the known scientific behaviors and then projecting these known behaviors on the data to look for deviations. When utilizing machine learning, this presents a particular challenge since we require that the model not only understands scientific data perfectly but also recognizes when the data is inconsistent and out of the scope of its trained behavior. In this paper, we present three datasets aimed at developing machine learning-based anomaly detection for disparate scientific domains covering astrophysics, genomics, and polar science. We present the different datasets along with a scheme to make machine learning challenges around the three datasets findable, accessible, interoperable, and reusable (FAIR). Furthermore, we present an approach that generalizes to future machine learning challenges, enabling the possibility of large, more compute-intensive challenges that can ultimately lead to scientific discovery.

Zahid Hassan Tushar, Adeleke Ademakinwa, Jianwu Wang et al.

Accurate cloud property retrieval is vital for understanding cloud behavior and its impact on climate, including applications in weather forecasting, climate modeling, and estimating Earth's radiation balance. The Independent Pixel Approximation (IPA), a widely used physics-based approach, simplifies radiative transfer calculations by assuming each pixel is independent of its neighbors. While computationally efficient, IPA has significant limitations, such as inaccuracies from 3D radiative effects, errors at cloud edges, and ineffectiveness for overlapping or heterogeneous cloud fields. Recent AI/ML-based deep learning models have improved retrieval accuracy by leveraging spatial relationships across pixels. However, these models are often memory-intensive, retrieve only a single cloud property, or struggle with joint property retrievals. To overcome these challenges, we introduce CloudUNet with Attention Module (CAM), a compact UNet-based model that employs attention mechanisms to reduce errors in thick, overlapping cloud regions and a specialized loss function for joint retrieval of Cloud Optical Thickness (COT) and Cloud Effective Radius (CER). Experiments on a Large Eddy Simulation (LES) dataset show that our CAM model outperforms state-of-the-art deep learning methods, reducing mean absolute errors (MAE) by 34% for COT and 42% for CER, and achieving 76% and 86% lower MAE for COT and CER retrievals compared to the IPA method.

Rohan Putatunda, Sanjay Purushotham, Ratnaksha Lele et al.

Drifting icebergs in the polar oceans play a key role in the Earth's climate system, impacting freshwater fluxes into the ocean and regional ecosystems while also posing a challenge to polar navigation. However, accurately forecasting iceberg trajectories remains a formidable challenge, primarily due to the scarcity of spatiotemporal data and the complex, nonlinear nature of iceberg motion, which is also impacted by environmental variables. The iceberg motion is influenced by multiple dynamic environmental factors, creating a highly variable system that makes trajectory identification complex. These limitations hinder the ability of deep learning models to effectively capture the underlying dynamics and provide reliable predictive outcomes. To address these challenges, we propose a hybrid IDRIFTNET model, a physics-driven deep learning model that combines an analytical formulation of iceberg drift physics, with an augmented residual learning model. The model learns the pattern of mismatch between the analytical solution and ground-truth observations, which is combined with a rotate-augmented spectral neural network that captures both global and local patterns from the data to forecast future iceberg drift positions. We compare IDRIFTNET model performance with state-of-the-art models on two Antarctic icebergs: A23A and B22A. Our findings demonstrate that IDRIFTNET outperforms other models by achieving a lower Final Displacement Error (FDE) and Average Displacement Error (ADE) across a variety of time points. These results highlight IDRIFTNET's effectiveness in capturing the complex, nonlinear drift of icebergs for forecasting iceberg trajectories under limited data and dynamic environmental conditions.

Zahid Hassan Tushar, Adeleke Ademakinwa, Jianwu Wang et al.

Cloud Optical Thickness (COT) is a critical cloud property influencing Earth's climate, weather, and radiation budget. Satellite radiance measurements enable global COT retrieval, but challenges like 3D cloud effects, viewing angles, and atmospheric interference must be addressed to ensure accurate estimation. Traditionally, the Independent Pixel Approximation (IPA) method, which treats individual pixels independently, has been used for COT estimation. However, IPA introduces significant bias due to its simplified assumptions. Recently, deep learning-based models have shown improved performance over IPA but lack robustness, as they are sensitive to variations in radiance intensity, distortions, and cloud shadows. These models also introduce substantial errors in COT estimation under different solar and viewing zenith angles. To address these challenges, we propose a novel angle-invariant, attention-based deep model called Cloud-Attention-Net with Angle Coding (CAAC). Our model leverages attention mechanisms and angle embeddings to account for satellite viewing geometry and 3D radiative transfer effects, enabling more accurate retrieval of COT. Additionally, our multi-angle training strategy ensures angle invariance. Through comprehensive experiments, we demonstrate that CAAC significantly outperforms existing state-of-the-art deep learning models, reducing cloud property retrieval errors by at least a factor of nine.

Bayu Adhi Tama, Vandana Janeja, Sanjay Purushotham

The increasing threat of sea level rise due to climate change necessitates a deeper understanding of ice sheet structures. This study addresses the need for accurate ice sheet data interpretation by introducing a suite of quantitative metrics designed to validate ice sheet annotation techniques. Focusing on both manual and automated methods, including ARESELP and its modified version, MARESELP, we assess their accuracy against expert annotations. Our methodology incorporates several computer vision metrics, traditionally underutilized in glaciological research, to evaluate the continuity and connectivity of ice layer annotations. The results demonstrate that while manual annotations provide invaluable expert insights, automated methods, particularly MARESELP, improve layer continuity and alignment with expert labels.

Catherine Ordun, Edward Raff, Sanjay Purushotham

Thermal images reveal medically important physiological information about human stress, signs of inflammation, and emotional mood that cannot be seen on visible images. Providing a method to generate thermal faces from visible images would be highly valuable for the telemedicine community in order to show this medical information. To the best of our knowledge, there are limited works on visible-to-thermal (VT) face translation, and many current works go the opposite direction to generate visible faces from thermal surveillance images (TV) for law enforcement applications. As a result, we introduce favtGAN, a VT GAN which uses the pix2pix image translation model with an auxiliary sensor label prediction network for generating thermal faces from visible images. Since most TV methods are trained on only one data source drawn from one thermal sensor, we combine datasets from faces and cityscapes. These combined data are captured from similar sensors in order to bootstrap the training and transfer learning task, especially valuable because visible-thermal face datasets are limited. Experiments on these combined datasets show that favtGAN demonstrates an increase in SSIM and PSNR scores of generated thermal faces, compared to training on a single face dataset alone.

Catherine Ordun, Edward Raff, Sanjay Purushotham

With the increased attention on thermal imagery for Covid-19 screening, the public sector may believe there are new opportunities to exploit thermal as a modality for computer vision and AI. Thermal physiology research has been ongoing since the late nineties. This research lies at the intersections of medicine, psychology, machine learning, optics, and affective computing. We will review the known factors of thermal vs. RGB imaging for facial emotion recognition. But we also propose that thermal imagery may provide a semi-anonymous modality for computer vision, over RGB, which has been plagued by misuse in facial recognition. However, the transition to adopting thermal imagery as a source for any human-centered AI task is not easy and relies on the availability of high fidelity data sources across multiple demographics and thorough validation. This paper takes the reader on a short review of machine learning in thermal FER and the limitations of collecting and developing thermal FER data for AI training. Our motivation is to provide an introductory overview into recent advances for thermal FER and stimulate conversation about the limitations in current datasets.

Catherine Ordun, Sanjay Purushotham, Edward Raff

This paper illustrates five different techniques to assess the distinctiveness of topics, key terms and features, speed of information dissemination, and network behaviors for Covid19 tweets. First, we use pattern matching and second, topic modeling through Latent Dirichlet Allocation (LDA) to generate twenty different topics that discuss case spread, healthcare workers, and personal protective equipment (PPE). One topic specific to U.S. cases would start to uptick immediately after live White House Coronavirus Task Force briefings, implying that many Twitter users are paying attention to government announcements. We contribute machine learning methods not previously reported in the Covid19 Twitter literature. This includes our third method, Uniform Manifold Approximation and Projection (UMAP), that identifies unique clustering-behavior of distinct topics to improve our understanding of important themes in the corpus and help assess the quality of generated topics. Fourth, we calculated retweeting times to understand how fast information about Covid19 propagates on Twitter. Our analysis indicates that the median retweeting time of Covid19 for a sample corpus in March 2020 was 2.87 hours, approximately 50 minutes faster than repostings from Chinese social media about H7N9 in March 2013. Lastly, we sought to understand retweet cascades, by visualizing the connections of users over time from fast to slow retweeting. As the time to retweet increases, the density of connections also increase where in our sample, we found distinct users dominating the attention of Covid19 retweeters. One of the simplest highlights of this analysis is that early-stage descriptive methods like regular expressions can successfully identify high-level themes which were consistently verified as important through every subsequent analysis.

Sanjay Purushotham, Chuizheng Meng, Zhengping Che et al.

Deep learning models (aka Deep Neural Networks) have revolutionized many fields including computer vision, natural language processing, speech recognition, and is being increasingly used in clinical healthcare applications. However, few works exist which have benchmarked the performance of the deep learning models with respect to the state-of-the-art machine learning models and prognostic scoring systems on publicly available healthcare datasets. In this paper, we present the benchmarking results for several clinical prediction tasks such as mortality prediction, length of stay prediction, and ICD-9 code group prediction using Deep Learning models, ensemble of machine learning models (Super Learner algorithm), SAPS II and SOFA scores. We used the Medical Information Mart for Intensive Care III (MIMIC-III) (v1.4) publicly available dataset, which includes all patients admitted to an ICU at the Beth Israel Deaconess Medical Center from 2001 to 2012, for the benchmarking tasks. Our results show that deep learning models consistently outperform all the other approaches especially when the `raw' clinical time series data is used as input features to the models.

Minh Nguyen, Sanjay Purushotham, Hien To et al.

Multivariate time series (MTS) have become increasingly common in healthcare domains where human vital signs and laboratory results are collected for predictive diagnosis. Recently, there have been increasing efforts to visualize healthcare MTS data based on star charts or parallel coordinates. However, such techniques might not be ideal for visualizing a large MTS dataset, since it is difficult to obtain insights or interpretations due to the inherent high dimensionality of MTS. In this paper, we propose 'm-TSNE': a simple and novel framework to visualize high-dimensional MTS data by projecting them into a low-dimensional (2-D or 3-D) space while capturing the underlying data properties. Our framework is easy to use and provides interpretable insights for healthcare professionals to understand MTS data. We evaluate our visualization framework on two real-world datasets and demonstrate that the results of our m-TSNE show patterns that are easy to understand while the other methods' visualization may have limitations in interpretability.

Zhengping Che, Sanjay Purushotham, Kyunghyun Cho et al.

Multivariate time series data in practical applications, such as health care, geoscience, and biology, are characterized by a variety of missing values. In time series prediction and other related tasks, it has been noted that missing values and their missing patterns are often correlated with the target labels, a.k.a., informative missingness. There is very limited work on exploiting the missing patterns for effective imputation and improving prediction performance. In this paper, we develop novel deep learning models, namely GRU-D, as one of the early attempts. GRU-D is based on Gated Recurrent Unit (GRU), a state-of-the-art recurrent neural network. It takes two representations of missing patterns, i.e., masking and time interval, and effectively incorporates them into a deep model architecture so that it not only captures the long-term temporal dependencies in time series, but also utilizes the missing patterns to achieve better prediction results. Experiments of time series classification tasks on real-world clinical datasets (MIMIC-III, PhysioNet) and synthetic datasets demonstrate that our models achieve state-of-the-art performance and provides useful insights for better understanding and utilization of missing values in time series analysis.

Shangwen Li, Sanjay Purushotham, Chen Chen et al.

Textual data such as tags, sentence descriptions are combined with visual cues to reduce the semantic gap for image retrieval applications in today's Multimodal Image Retrieval (MIR) systems. However, all tags are treated as equally important in these systems, which may result in misalignment between visual and textual modalities during MIR training. This will further lead to degenerated retrieval performance at query time. To address this issue, we investigate the problem of tag importance prediction, where the goal is to automatically predict the tag importance and use it in image retrieval. To achieve this, we first propose a method to measure the relative importance of object and scene tags from image sentence descriptions. Using this as the ground truth, we present a tag importance prediction model to jointly exploit visual, semantic and context cues. The Structural Support Vector Machine (SSVM) formulation is adopted to ensure efficient training of the prediction model. Then, the Canonical Correlation Analysis (CCA) is employed to learn the relation between the image visual feature and tag importance to obtain robust retrieval performance. Experimental results on three real-world datasets show a significant performance improvement of the proposed MIR with Tag Importance Prediction (MIR/TIP) system over other MIR systems.

Zhengping Che, Sanjay Purushotham, Robinder Khemani et al.

Exponential growth in Electronic Healthcare Records (EHR) has resulted in new opportunities and urgent needs for discovery of meaningful data-driven representations and patterns of diseases in Computational Phenotyping research. Deep Learning models have shown superior performance for robust prediction in computational phenotyping tasks, but suffer from the issue of model interpretability which is crucial for clinicians involved in decision-making. In this paper, we introduce a novel knowledge-distillation approach called Interpretable Mimic Learning, to learn interpretable phenotype features for making robust prediction while mimicking the performance of deep learning models. Our framework uses Gradient Boosting Trees to learn interpretable features from deep learning models such as Stacked Denoising Autoencoder and Long Short-Term Memory. Exhaustive experiments on a real-world clinical time-series dataset show that our method obtains similar or better performance than the deep learning models, and it provides interpretable phenotypes for clinical decision making.

Sanjay Purushotham, Yan Liu, C. -C. Jay Kuo

Social network websites, such as Facebook, YouTube, Lastfm etc, have become a popular platform for users to connect with each other and share content or opinions. They provide rich information for us to study the influence of user's social circle in their decision process. In this paper, we are interested in examining the effectiveness of social network information to predict the user's ratings of items. We propose a novel hierarchical Bayesian model which jointly incorporates topic modeling and probabilistic matrix factorization of social networks. A major advantage of our model is to automatically infer useful latent topics and social information as well as their importance to collaborative filtering from the training data. Empirical experiments on two large-scale datasets show that our algorithm provides a more effective recommendation system than the state-of-the art approaches. Our results reveal interesting insight that the social circles have more influence on people's decisions about the usefulness of information (e.g., bookmarking preference on Delicious) than personal taste (e.g., music preference on Lastfm). We also examine and discuss solutions on potential information leak in many recommendation systems that utilize social information.