Medical machine learning algorithms are typically evaluated based on accuracy vs. a clinician-defined ground truth, a reasonable initial choice since trained clinicians are usually better classifiers than ML models. However, this metric does not fully capture the actual clinical task: it neglects the fact that humans, even with perfect accuracy, are subject to non-trivial error from the Poisson statistics of rare events, because clinical protocols often specify a relatively small sample size. For example, to quantitate malaria on a thin blood film a clinician examines only 2000 red blood cells (0.0004 uL), which can yield large Poisson variation in the actual number of parasites present, so that a perfect human's count can differ substantially from the true average load. In contrast, an ML system may be less accurate on an object level, but it may also have the option to examine more blood (e.g. 0.1 uL, or 250x). Then while its parasite identification error is higher, the Poisson variability of its estimate is lower due to larger sample size. To qualify for clinical deployment, an ML system's performance must match current standard of care, typically a very demanding target. To achieve this, it may be possible to offset the ML system's lower accuracy by increasing its sample size to reduce Poisson error, and thus attain the same net clinical performance as a perfectly accurate human limited by smaller sample size. In this paper, we analyse the mathematics of the relationship between Poisson error, classification error, and total error. This mathematical toolkit enables teams optimizing ML systems to leverage a relative strength (larger sample sizes) to offset a relative weakness (classification accuracy). We illustrate the methods with two concrete examples: diagnosis and quantitation of malaria on blood films.

Conceptualization, or viewing entities and situations as instances of abstract concepts in mind and making inferences based on that, is a vital component in human intelligence for commonsense reasoning. Despite recent progress in artificial intelligence to acquire and model commonsense attributed to neural language models and commonsense knowledge graphs (CKGs), conceptualization is yet to be introduced thoroughly, making current approaches ineffective to cover knowledge about countless diverse entities and situations in the real world. To address the problem, we thoroughly study the role of conceptualization in commonsense reasoning, and formulate a framework to replicate human conceptual induction by acquiring abstract knowledge about events regarding abstract concepts, as well as higher-level triples or inferences upon them. We then apply the framework to ATOMIC, a large-scale human-annotated CKG, aided by the taxonomy Probase. We annotate a dataset on the validity of contextualized conceptualizations from ATOMIC on both event and triple levels, develop a series of heuristic rules based on linguistic features, and train a set of neural models to generate and verify abstract knowledge. Based on these components, a pipeline to acquire abstract knowledge is built. A large abstract CKG upon ATOMIC is then induced, ready to be instantiated to infer about unseen entities or situations. Finally, we empirically show the benefits of augmenting CKGs with abstract knowledge in downstream tasks like commonsense inference and zero-shot commonsense QA.

Big data has ushered in a new wave of predictive power using machine learning models. In this work, we assess what {\it big} means in the context of typical materials-science machine-learning problems. This concerns not only data volume, but also data quality and veracity as much as infrastructure issues. With selected examples, we ask (i) how models generalize to similar datasets, (ii) how high-quality datasets can be gathered from heterogenous sources, (iii) how the feature set and complexity of a model can affect expressivity, and (iv) what infrastructure requirements are needed to create larger datasets and train models on them. In sum, we find that big data present unique challenges along very different aspects that should serve to motivate further work.

Since the inception of the SARS - CoV - 2 (COVID - 19) novel coronavirus, a lot of time and effort is being allocated to estimate the trajectory and possibly, forecast with a reasonable degree of accuracy, the number of cases, recoveries, and deaths due to the same. The model proposed in this paper is a mindful step in the same direction. The primary model in question is a Hybrid Holt's Model embedded with a Wavelet-based ANN. To test its forecasting ability, we have compared three separate models, the first, being a simple ARIMA model, the second, also an ARIMA model with a wavelet-based function, and the third, being the proposed model. We have also compared the forecast accuracy of this model with that of a modern day Vanilla LSTM recurrent neural network model. We have tested the proposed model on the number of confirmed cases (daily) for the entire country as well as 6 hotspot states. We have also proposed a simple adjustment algorithm in addition to the hybrid model so that daily and/or weekly forecasts can be meted out, with respect to the entirety of the country, as well as a moving window performance metric based on out-of-sample forecasts. In order to have a more rounded approach to the analysis of COVID-19 dynamics, focus has also been given to the estimation of the Basic Reproduction Number, $R_0$ using a compartmental epidemiological model (SIR). Lastly, we have also given substantial attention to estimating the shelf-life of the proposed model. It is obvious yet noteworthy how an accurate model, in this regard, can ensure better allocation of healthcare resources, as well as, enable the government to take necessary measures ahead of time.

The COVID-19 Forecast Hub was founded in 2020 and serves as a "central repository of COVID-19 forecasts from over 50 independent research groups" [1]. Participant research groups submit county, state and national US COVID-19 forecasts with a standardized format; and the Forecast Hub provides an interactive visualization tool to help decision makers and the general public analyze weekly predictions for COVID-19 hospitalizations, cases and deaths. The standardized predictions collected from all research groups, as well as the predictions for an ensemble model that brings all individual predictions together, are also shared with the Centers for Disease Control and Prevention (CDC) who uses these results for their official COVID-19 communications [2]. The COVID-19 Forecast Hub has been, and continues to be, a critical centralized resource to promote transparent decision making. Nevertheless, by focusing exclusively on prediction accuracy at different spatial granularities (e.g., county or state), the Forecast Hub fails to evaluate whether the proposed models are fair i.e., share similar prediction performance across social determinants that have been known to play a role in COVID-19 including race, ethnicity and rurality [3, 4]. Diverse prediction performance across social determinants - for example, higher prediction errors for a given minority race or ethnicity - could negatively impact resource allocation and intervention decisions e.g., hospital beds or stay-at-home orders, given that the CDC appears to be using the Forecast Hub predictions for official communications that subsequently inform policy decisions [2]. In other words, allocation or intervention harms might occur if models from the Forecast Hub are used to inform decision making across communities without taking into account fairness metrics [5]. There are many reasons why the COVID-19 prediction performance can be different across social determinants such as race, ethnicity or urbanization levels. The Forecast Hub's COVID-19 prediction models are trained on datasets containing COVID-19

Subtitling plays a crucial role in enhancing the accessibility of audiovisual content and encompasses three primary subtasks: translating spoken dialogue, segmenting translations into concise textual units, and estimating timestamps that govern their on-screen duration. Past attempts to automate this process rely, to varying degrees, on automatic transcripts, employed diversely for the three subtasks. In response to the acknowledged limitations associated with this reliance on transcripts, recent research has shifted towards transcription-free solutions for translation and segmentation, leaving the direct generation of timestamps as uncharted territory. To fill this gap, we introduce the first direct model capable of producing automatic subtitles, entirely eliminating any dependence on intermediate transcripts also for timestamp prediction. Experimental results, backed by manual evaluation, showcase our solution's new state-of-the-art performance across multiple language pairs and diverse conditions.

Large language models (LLMs) have taken centre stage in debates on Artificial Intelligence. Yet there remains a gap in how to assess LLMs' conformity to important human values. In this paper, we investigate whether state-of-the-art LLMs, GPT-4 and Claude 2.1 (Gemini Pro and LLAMA 2 did not generate valid results) are moral hypocrites. We employ two research instruments based on the Moral Foundations Theory: (i) the Moral Foundations Questionnaire (MFQ), which investigates which values are considered morally relevant in abstract moral judgements; and (ii) the Moral Foundations Vignettes (MFVs), which evaluate moral cognition in concrete scenarios related to each moral foundation. We characterise conflicts in values between these different abstractions of moral evaluation as hypocrisy. We found that both models displayed reasonable consistency within each instrument compared to humans, but they displayed contradictory and hypocritical behaviour when we compared the abstract values present in the MFQ to the evaluation of concrete moral violations of the MFV.

The rapid development of Large Language Models (LLMs) demonstrates remarkable multilingual capabilities in natural language processing, attracting global attention in both academia and industry. To mitigate potential discrimination and enhance the overall usability and accessibility for diverse language user groups, it is important for the development of language-fair technology. Despite the breakthroughs of LLMs, the investigation into the multilingual scenario remains insufficient, where a comprehensive survey to summarize recent approaches, developments, limitations, and potential solutions is desirable. To this end, we provide a survey with multiple perspectives on the utilization of LLMs in the multilingual scenario. We first rethink the transitions between previous and current research on pre-trained language models. Then we introduce several perspectives on the multilingualism of LLMs, including training and inference methods, model security, multi-domain with language culture, and usage of datasets. We also discuss the major challenges that arise in these aspects, along with possible solutions. Besides, we highlight future research directions that aim at further enhancing LLMs with multilingualism. The survey aims to help the research community address multilingual problems and provide a comprehensive understanding of the core concepts, key techniques, and latest developments in multilingual natural language processing based on LLMs.

In this paper, we introduce an alternative approach to enhancing Multi-Agent Reinforcement Learning (MARL) through the integration of domain knowledge and attention-based policy mechanisms. Our methodology focuses on the incorporation of domain-specific expertise into the learning process, which simplifies the development of collaborative behaviors. This approach aims to reduce the complexity and learning overhead typically associated with MARL by enabling agents to concentrate on essential aspects of complex tasks, thus optimizing the learning curve. The utilization of attention mechanisms plays a key role in our model. It allows for the effective processing of dynamic context data and nuanced agent interactions, leading to more refined decision-making. Applied in standard MARL scenarios, such as the Stanford Intelligent Systems Laboratory (SISL) Pursuit and Multi-Particle Environments (MPE) Simple Spread, our method has been shown to improve both learning efficiency and the effectiveness of collaborative behaviors. The results indicate that our attention-based approach can be a viable approach for improving the efficiency of MARL training process, integrating domain-specific knowledge at the action level.

The field of natural language processing (NLP) has recently witnessed a transformative shift with the emergence of foundation models, particularly Large Language Models (LLMs) that have revolutionized text-based NLP. This paradigm has extended to other modalities, including speech, where researchers are actively exploring the combination of Speech Foundation Models (SFMs) and LLMs into single, unified models capable of addressing multimodal tasks. Among such tasks, this paper focuses on speech-to-text translation (ST). By examining the published papers on the topic, we propose a unified view of the architectural solutions and training strategies presented so far, highlighting similarities and differences among them. Based on this examination, we not only organize the lessons learned but also show how diverse settings and evaluation approaches hinder the identification of the best-performing solution for each architectural building block and training choice. Lastly, we outline recommendations for future works on the topic aimed at better understanding the strengths and weaknesses of the SFM+LLM solutions for ST.