Healthcare data in the United States often records only a patient's coarse race group: for example, both Indian and Chinese patients are typically coded as "Asian." It is unknown, however, whether this coarse coding conceals meaningful disparities in the performance of clinical risk scores across granular race groups. Here we show that it does. Using data from 418K emergency department visits, we assess clinical risk score performance disparities across 26 granular groups for three outcomes, five risk scores, and four performance metrics. Across outcomes and metrics, we show that the risk scores exhibit significant granular performance disparities within coarse race groups. In fact, variation in performance within coarse groups often *exceeds* the variation between coarse groups. We explore why these disparities arise, finding that outcome rates, feature distributions, and the relationships between features and outcomes all vary significantly across granular groups. Our results suggest that healthcare providers, hospital systems, and machine learning researchers should strive to collect, release, and use granular race data in place of coarse race data, and that existing analyses may significantly underestimate racial disparities in performance.

Predictive algorithms inform consequential decisions in settings where the outcome is selectively observed given some choices made by human decision makers. There often exists unobserved confounders that affected the decision maker's choice and the outcome. We propose a unified methodology for the robust design and evaluation of predictive algorithms in selectively observed data under such unobserved confounding. Our approach imposes general assumptions on how much the outcome may vary on average between unselected and selected units conditional on observed covariates and identified nuisance parameters, formalizing popular empirical strategies for imputing missing data such as proxy outcomes and instrumental variables. We develop debiased machine learning estimators for the bounds on a large class of predictive performance estimands, such as the conditional likelihood of the outcome, a predictive algorithm's mean square error, true/false positive rate, and many others, under these assumptions. In an administrative dataset from a large Australian financial institution, we illustrate how varying assumptions on unobserved confounding leads to meaningful changes in default risk predictions and evaluations of credit scores across sensitive groups.

The automation of a large number of software engineering tasks is becoming possible thanks to Machine Learning (ML). Central to applying ML to software artifacts (like source or executable code) is converting them into forms suitable for learning. Traditionally, researchers have relied on manually selected features, based on expert knowledge which is sometimes imprecise and generally incomplete. Representation learning has allowed ML to automatically choose suitable representations and relevant features. Yet, for Android-related tasks, existing models like apk2vec focus on whole-app levels, or target specific tasks like smali2vec, which limits their applicability. Our work is part of a new line of research that investigates effective, task-agnostic, and fine-grained universal representations of bytecode to mitigate both of these two limitations. Such representations aim to capture information relevant to various low-level downstream tasks (e.g., at the class-level). We are inspired by the field of Natural Language Processing, where the problem of universal representation was addressed by building Universal Language Models, such as BERT, whose goal is to capture abstract semantic information about sentences, in a way that is reusable for a variety of tasks. We propose DexBERT, a BERT-like Language Model dedicated to representing chunks of DEX bytecode, the main binary format used in Android applications. We empirically assess whether DexBERT is able to model the DEX language and evaluate the suitability of our model in three distinct class-level software engineering tasks: Malicious Code Localization, Defect Prediction, and Component Type Classification. We also experiment with strategies to deal with the problem of catering to apps having vastly different sizes, and we demonstrate one example of using our technique to investigate what information is relevant to a given task.

White matter bundle segmentation is a cornerstone of modern tractography to study the brain's structural connectivity in domains such as neurological disorders, neurosurgery, and aging. In this study, we present FIESTA (FIbEr Segmentation in Tractography using Autoencoders), a reliable and robust, fully automated, and easily semi-automatically calibrated pipeline based on deep autoencoders that can dissect and fully populate white matter bundles. This pipeline is built upon previous works that demonstrated how autoencoders can be used successfully for streamline filtering, bundle segmentation, and streamline generation in tractography. Our proposed method improves bundle segmentation coverage by recovering hard-to-track bundles with generative sampling through the latent space seeding of the subject bundle and the atlas bundle. A latent space of streamlines is learned using autoencoder-based modeling combined with contrastive learning. Using an atlas of bundles in standard space (MNI), our proposed method segments new tractograms using the autoencoder latent distance between each tractogram streamline and its closest neighbor bundle in the atlas of bundles. Intra-subject bundle reliability is improved by recovering hard-to-track streamlines, using the autoencoder to generate new streamlines that increase the spatial coverage of each bundle while remaining anatomically correct. Results show that our method is more reliable than state-of-the-art automated virtual dissection methods such as RecoBundles, RecoBundlesX, TractSeg, White Matter Analysis and XTRACT. Our framework allows for the transition from one anatomical bundle definition to another with marginal calibration efforts. Overall, these results show that our framework improves the practicality and usability of current state-of-the-art bundle segmentation framework.

Sign language, which conveys meaning through gestures, is the chief means of communication among deaf people. Recognizing sign language in natural settings presents significant challenges due to factors such as lighting, background clutter, and variations in signer characteristics. In this thesis, I study automatic sign language processing in the wild, using signing videos collected from the Internet. This thesis contributes new datasets, tasks, and methods. Most chapters of this thesis address tasks related to fingerspelling, an important component of sign language and yet has not been studied widely by prior work. I present three new large-scale ASL datasets in the wild: ChicagoFSWild, ChicagoFSWild+, and OpenASL. Using ChicagoFSWild and ChicagoFSWild+, I address fingerspelling recognition, which consists of transcribing fingerspelling sequences into text. I propose an end-to-end approach based on iterative attention that allows recognition from a raw video without explicit hand detection. I further show that using a Conformer-based network jointly modeling handshape and mouthing can bring performance close to that of humans. Next, I propose two tasks for building real-world fingerspelling-based applications: fingerspelling detection and search. For fingerspelling detection, I introduce a suite of evaluation metrics and a new detection model via multi-task training. To address the problem of searching for fingerspelled keywords in raw sign language videos, we propose a novel method that jointly localizes and matches fingerspelling segments to text. Finally, I will describe a benchmark for large-vocabulary open-domain sign language translation based on OpenASL. To address the challenges of sign language translation in realistic settings, we propose a set of techniques including sign search as a pretext task for pre-training and fusion of mouthing and handshape features.

We examine the disconnect between scholarship and practice in applying machine learning to trust and safety problems, using misinformation detection as a case study. We systematize literature on automated detection of misinformation across a corpus of 270 well-cited papers in the field. We then examine subsets of papers for data and code availability, design missteps, reproducibility, and generalizability. We find significant shortcomings in the literature that call into question claimed performance and practicality. Detection tasks are often meaningfully distinct from the challenges that online services actually face. Datasets and model evaluation are often non-representative of real-world contexts, and evaluation frequently is not independent of model training. Data and code availability is poor. Models do not generalize well to out-of-domain data. Based on these results, we offer recommendations for evaluating machine learning applications to trust and safety problems. Our aim is for future work to avoid the pitfalls that we identify.

In a connection of many IoT devices that each collect data, normally training a machine learning model would involve transmitting the data to a central server which requires strict privacy rules. However, some owners are reluctant of availing their data out of the company due to data security concerns. Federated learning(FL) as a distributed machine learning approach performs training of a machine learning model on the device that gathered the data itself. In this scenario, data is not share over the network for training purpose. Fedavg as one of FL algorithms permits a model to be copied to participating devices during a training session. The devices could be chosen at random, and a device can be aborted. The resulting models are sent to the coordinating server and then average models from the devices that finished training. The process is repeated until a desired model accuracy is achieved. By doing this, FL approach solves the privacy problem for IoT/ IIoT devices that held sensitive data for the owners. In this paper, we leverage the benefits of FL and implemented Fedavg algorithm on a recent dataset that represent the modern IoT/ IIoT device networks. The results were almost the same as the centralized machine learning approach. We also evaluated some shortcomings of Fedavg such as unfairness that happens during the training when struggling devices do not participate for every stage of training. This inefficient training of local or global model could lead in a high number of false alarms in intrusion detection systems for IoT/IIoT gadgets developed using Fedavg. Hence, after evaluating the FedAv deep auto encoder with centralized deep auto encoder ML, we further proposed and designed a Fair Fedavg algorithm that will be evaluated in the future work.

Prostate cancer pathology plays a crucial role in clinical management but is time-consuming. Artificial intelligence (AI) shows promise in detecting prostate cancer and grading patterns. We tested an AI-based digital twin of a pathologist, vPatho, on 2,603 histology images of prostate tissue stained with hematoxylin and eosin. We analyzed various factors influencing tumor-grade disagreement between vPatho and six human pathologists. Our results demonstrated that vPatho achieved comparable performance in prostate cancer detection and tumor volume estimation, as reported in the literature. Concordance levels between vPatho and human pathologists were examined. Notably, moderate to substantial agreement was observed in identifying complementary histological features such as ductal, cribriform, nerve, blood vessels, and lymph cell infiltrations. However, concordance in tumor grading showed a decline when applied to prostatectomy specimens (kappa = 0.44) compared to biopsy cores (kappa = 0.70). Adjusting the decision threshold for the secondary Gleason pattern from 5% to 10% improved the concordance level between pathologists and vPatho for tumor grading on prostatectomy specimens (kappa from 0.44 to 0.64). Potential causes of grade discordance included the vertical extent of tumors toward the prostate boundary and the proportions of slides with prostate cancer. Gleason pattern 4 was particularly associated with discordance. Notably, grade discordance with vPatho was not specific to any of the six pathologists involved in routine clinical grading. In conclusion, our study highlights the potential utility of AI in developing a digital twin of a pathologist. This approach can help uncover limitations in AI adoption and the current grading system for prostate cancer pathology.

Multi-modal systems have the capacity of producing more reliable results than systems with a single modality in road detection due to perceiving different aspects of the scene. We focus on using raw sensor inputs instead of, as it is typically done in many SOTA works, leveraging architectures that require high pre-processing costs such as surface normals or dense depth predictions. By using raw sensor inputs, we aim to utilize a low-cost model thatminimizes both the pre-processing andmodel computation costs. This study presents a cost-effective and highly accurate solution for road segmentation by integrating data from multiple sensorswithin a multi-task learning architecture.Afusion architecture is proposed in which RGB and LiDAR depth images constitute the inputs of the network. Another contribution of this study is to use IMU/GNSS (inertial measurement unit/global navigation satellite system) inertial navigation system whose data is collected synchronously and calibrated with a LiDAR-camera to compute aggregated dense LiDAR depth images. It has been demonstrated by experiments on the KITTI dataset that the proposed method offers fast and high-performance solutions. We have also shown the performance of our method on Cityscapes where raw LiDAR data is not available. The segmentation results obtained for both full and half resolution images are competitive with existing methods. Therefore, we conclude that our method is not dependent only on raw LiDAR data; rather, it can be used with different sensor modalities. The inference times obtained in all experiments are very promising for real-time experiments.

Out-of-distribution (OOD) detection plays a vital role in enhancing the reliability of machine learning (ML) models. The emergence of large language models (LLMs) has catalyzed a paradigm shift within the ML community, showcasing their exceptional capabilities across diverse natural language processing tasks. While existing research has probed OOD detection with relative small-scale Transformers like BERT, RoBERTa and GPT-2, the stark differences in scales, pre-training objectives, and inference paradigms call into question the applicability of these findings to LLMs. This paper embarks on a pioneering empirical investigation of OOD detection in the domain of LLMs, focusing on LLaMA series ranging from 7B to 65B in size. We thoroughly evaluate commonly-used OOD detectors, scrutinizing their performance in both zero-grad and fine-tuning scenarios. Notably, we alter previous discriminative in-distribution fine-tuning into generative fine-tuning, aligning the pre-training objective of LLMs with downstream tasks. Our findings unveil that a simple cosine distance OOD detector demonstrates superior efficacy, outperforming other OOD detectors. We provide an intriguing explanation for this phenomenon by highlighting the isotropic nature of the embedding spaces of LLMs, which distinctly contrasts with the anisotropic property observed in smaller BERT family models. The new insight enhances our understanding of how LLMs detect OOD data, thereby enhancing their adaptability and reliability in dynamic environments.