Alerting the public when heat may harm their health is a crucial service, especially considering that extreme heat events will be more frequent under climate change. Current practice for issuing heat alerts in the US does not take advantage of modern data science methods for optimizing local alert criteria. Specifically, application of reinforcement learning (RL) has the potential to inform more health-protective policies, accounting for regional and sociodemographic heterogeneity as well as sequential dependence of alerts. In this work, we formulate the issuance of heat alerts as a sequential decision making problem and develop modifications to the RL workflow to address challenges commonly encountered in environmental health settings. Key modifications include creating a simulator that pairs hierarchical Bayesian modeling of low-signal health effects with sampling of real weather trajectories (exogenous features), constraining the total number of alerts issued as well as preventing alerts on less-hot days, and optimizing location-specific policies. Post-hoc contrastive analysis offers insights into scenarios when using RL for heat alert issuance may protect public health better than the current or alternative policies. This work contributes to a broader movement of advancing data-driven policy optimization for public health and climate change adaptation.
相關內容
In the rapidly evolving landscape of medical documentation, transcribing clinical dialogues accurately is increasingly paramount. This study explores the potential of Large Language Models (LLMs) to enhance the accuracy of Automatic Speech Recognition (ASR) systems in medical transcription. Utilizing the PriMock57 dataset, which encompasses a diverse range of primary care consultations, we apply advanced LLMs to refine ASR-generated transcripts. Our research is multifaceted, focusing on improvements in general Word Error Rate (WER), Medical Concept WER (MC-WER) for the accurate transcription of essential medical terms, and speaker diarization accuracy. Additionally, we assess the role of LLM post-processing in improving semantic textual similarity, thereby preserving the contextual integrity of clinical dialogues. Through a series of experiments, we compare the efficacy of zero-shot and Chain-of-Thought (CoT) prompting techniques in enhancing diarization and correction accuracy. Our findings demonstrate that LLMs, particularly through CoT prompting, not only improve the diarization accuracy of existing ASR systems but also achieve state-of-the-art performance in this domain. This improvement extends to more accurately capturing medical concepts and enhancing the overall semantic coherence of the transcribed dialogues. These findings illustrate the dual role of LLMs in augmenting ASR outputs and independently excelling in transcription tasks, holding significant promise for transforming medical ASR systems and leading to more accurate and reliable patient records in healthcare settings.
Large language models have the potential to be valuable in the healthcare industry, but it's crucial to verify their safety and effectiveness through rigorous evaluation. For this purpose, we comprehensively evaluated both open-source LLMs and Google's new multimodal LLM called Gemini across Medical reasoning, hallucination detection, and Medical Visual Question Answering tasks. While Gemini showed competence, it lagged behind state-of-the-art models like MedPaLM 2 and GPT-4 in diagnostic accuracy. Additionally, Gemini achieved an accuracy of 61.45\% on the medical VQA dataset, significantly lower than GPT-4V's score of 88\%. Our analysis revealed that Gemini is highly susceptible to hallucinations, overconfidence, and knowledge gaps, which indicate risks if deployed uncritically. We also performed a detailed analysis by medical subject and test type, providing actionable feedback for developers and clinicians. To mitigate risks, we applied prompting strategies that improved performance. Additionally, we facilitated future research and development by releasing a Python module for medical LLM evaluation and establishing a dedicated leaderboard on Hugging Face for medical domain LLMs. Python module can be found at //github.com/promptslab/RosettaEval
In order to optimize the radiotherapy delivery for cancer treatment, especially when dealing with complex treatments such as Total Marrow and Lymph Node Irradiation (TMLI), the accurate contouring of the Planning Target Volume (PTV) is crucial. Unfortunately, relying on manual contouring for such treatments is time-consuming and prone to errors. In this paper, we investigate the application of Deep Learning (DL) to automate the segmentation of the PTV in TMLI treatment, building upon previous work that introduced a solution to this problem based on a 2D U-Net model. We extend the previous research (i) by employing the nnU-Net framework to develop both 2D and 3D U-Net models and (ii) by evaluating the trained models on the PTV with the exclusion of bones, which consist mainly of lymp-nodes and represent the most challenging region of the target volume to segment. Our result show that the introduction of nnU-NET framework led to statistically significant improvement in the segmentation performance. In addition, the analysis on the PTV after the exclusion of bones showed that the models are quite robust also on the most challenging areas of the target volume. Overall, our study is a significant step forward in the application of DL in a complex radiotherapy treatment such as TMLI, offering a viable and scalable solution to increase the number of patients who can benefit from this treatment.
Affordances are fundamental descriptors of relationships between actions, objects and effects. They provide the means whereby a robot can predict effects, recognize actions, select objects and plan its behavior according to desired goals. This paper approaches the problem of an embodied agent exploring the world and learning these affordances autonomously from its sensory experiences. Models exist for learning the structure and the parameters of a Bayesian Network encoding this knowledge. Although Bayesian Networks are capable of dealing with uncertainty and redundancy, previous work considered complete observability of the discrete sensory data, which may lead to hard errors in the presence of noise. In this paper we consider a probabilistic representation of the sensors by Gaussian Mixture Models (GMMs) and explicitly taking into account the probability distribution contained in each discrete affordance concept, which can lead to a more correct learning.
Consider a panel data setting where repeated observations on individuals are available. Often it is reasonable to assume that there exist groups of individuals that share similar effects of observed characteristics, but the grouping is typically unknown in advance. We first conduct a local analysis which reveals that the variances of the individual coefficient estimates contain useful information for the estimation of group structure. We then propose a method to estimate unobserved groupings for general panel data models that explicitly account for the variance information. Our proposed method remains computationally feasible with a large number of individuals and/or repeated measurements on each individual. The developed ideas can also be applied even when individual-level data are not available and only parameter estimates together with some quantification of estimation uncertainty are given to the researcher. A thorough simulation study demonstrates superior performance of our method than existing methods and we apply the method to two empirical applications.
Face recognition (FR) has reached a high technical maturity. However, its use needs to be carefully assessed from an ethical perspective, especially in sensitive scenarios. This is precisely the focus of this paper: the use of FR for the identification of specific subjects in moderately to densely crowded spaces (e.g. public spaces, sports stadiums, train stations) and law enforcement scenarios. In particular, there is a need to consider the trade-off between the need to protect privacy and fundamental rights of citizens as well as their safety. Recent Artificial Intelligence (AI) policies, notably the European AI Act, propose that such FR interventions should be proportionate and deployed only when strictly necessary. Nevertheless, concrete guidelines on how to address the concept of proportional FR intervention are lacking to date. This paper proposes a framework to contribute to assessing whether an FR intervention is proportionate or not for a given context of use in the above mentioned scenarios. It also identifies the main quantitative and qualitative variables relevant to the FR intervention decision (e.g. number of people in the scene, level of harm that the person(s) in search could perpetrate, consequences to individual rights and freedoms) and propose a 2D graphical model making it possible to balance these variables in terms of ethical cost vs security gain. Finally, different FR scenarios inspired by real-world deployments validate the proposed model. The framework is conceived as a simple support tool for decision makers when confronted with the deployment of an FR system.
Loneliness is a growing health concern as it can lead to depression and other associated mental health problems for people who experience feelings of loneliness over prolonged periods of time. Utilizing passive sensing methods that use smartphone and wearable sensor data to capture daily behavioural patterns offers a promising approach for the early detection of loneliness. Given the subjective nature of loneliness and people's varying daily routines, past detection approaches using machine learning models often face challenges with effectively detecting loneliness. This paper proposes a methodologically novel approach, particularly developing a loneliness detection system that evolves over time, adapts to new data, and provides real-time detection. Our study utilized the Globem dataset, a comprehensive collection of passive sensing data acquired over 10 weeks from university students. The base of our approach is the continuous identification and refinement of similar behavioural groups among students using an incremental clustering method. As we add new data, the model improves based on changing behavioural patterns. Parallel to this, we create and update classification models to detect loneliness among the evolving behavioural groups of students. When unique behavioural patterns are observed among student data, specialized classification models have been created. For predictions of loneliness, a collaborative effort between the generalized and specialized models is employed, treating each prediction as a vote. This study's findings reveal that group-based loneliness detection models exhibit superior performance compared to generic models, underscoring the necessity for more personalized approaches tailored to specific behavioural patterns. These results pave the way for future research, emphasizing the development of finely-tuned, individualized mental health interventions.
Estimating Fold Changes from Partially Observed Outcomes with Applications in Microbial Metagenomics
We consider the problem of estimating fold-changes in the expected value of a multivariate outcome that is observed subject to unknown sample-specific and category-specific perturbations. We are motivated by high-throughput sequencing studies of the abundance of microbial taxa, in which microbes are systematically over- and under-detected relative to their true abundances. Our log-linear model admits a partially identifiable estimand, and we establish full identifiability by imposing interpretable parameter constraints. To reduce bias and guarantee the existence of parameter estimates in the presence of sparse observations, we apply an asymptotically negligible and constraint-invariant penalty to our estimating function. We develop a fast coordinate descent algorithm for estimation, and an augmented Lagrangian algorithm for estimation under null hypotheses. We construct a model-robust score test, and demonstrate valid inference even for small sample sizes and violated distributional assumptions. The flexibility of the approach and comparisons to related methods are illustrated via a meta-analysis of microbial associations with colorectal cancer.
Textual entailment is a fundamental task in natural language processing. Most approaches for solving the problem use only the textual content present in training data. A few approaches have shown that information from external knowledge sources like knowledge graphs (KGs) can add value, in addition to the textual content, by providing background knowledge that may be critical for a task. However, the proposed models do not fully exploit the information in the usually large and noisy KGs, and it is not clear how it can be effectively encoded to be useful for entailment. We present an approach that complements text-based entailment models with information from KGs by (1) using Personalized PageR- ank to generate contextual subgraphs with reduced noise and (2) encoding these subgraphs using graph convolutional networks to capture KG structure. Our technique extends the capability of text models exploiting structural and semantic information found in KGs. We evaluate our approach on multiple textual entailment datasets and show that the use of external knowledge helps improve prediction accuracy. This is particularly evident in the challenging BreakingNLI dataset, where we see an absolute improvement of 5-20% over multiple text-based entailment models.
Object detection typically assumes that training and test data are drawn from an identical distribution, which, however, does not always hold in practice. Such a distribution mismatch will lead to a significant performance drop. In this work, we aim to improve the cross-domain robustness of object detection. We tackle the domain shift on two levels: 1) the image-level shift, such as image style, illumination, etc, and 2) the instance-level shift, such as object appearance, size, etc. We build our approach based on the recent state-of-the-art Faster R-CNN model, and design two domain adaptation components, on image level and instance level, to reduce the domain discrepancy. The two domain adaptation components are based on H-divergence theory, and are implemented by learning a domain classifier in adversarial training manner. The domain classifiers on different levels are further reinforced with a consistency regularization to learn a domain-invariant region proposal network (RPN) in the Faster R-CNN model. We evaluate our newly proposed approach using multiple datasets including Cityscapes, KITTI, SIM10K, etc. The results demonstrate the effectiveness of our proposed approach for robust object detection in various domain shift scenarios.
小貼士
登錄享
注冊地址： 北京市海淀區羊坊店路18號2幢3層301-191