Identification of best performing fertilizer practices among a set of contrasting practices with field trials is challenging as crop losses are costly for farmers. To identify best management practices, an ''intuitive strategy'' would be to set multi-year field trials with equal proportion of each practice to test. Our objective was to provide an identification strategy using a bandit algorithm that was better at minimizing farmers' losses occurring during the identification, compared with the ''intuitive strategy''. We used a modification of the Decision Support Systems for Agro-Technological Transfer (DSSAT) crop model to mimic field trial responses, with a case-study in Southern Mali. We compared fertilizer practices using a risk-aware measure, the Conditional Value-at-Risk (CVaR), and a novel agronomic metric, the Yield Excess (YE). YE accounts for both grain yield and agronomic nitrogen use efficiency. The bandit-algorithm performed better than the intuitive strategy: it increased, in most cases, farmers' protection against worst outcomes. This study is a methodological step which opens up new horizons for risk-aware ensemble identification of the performance of contrasting crop management practices in real conditions.

Mental health research through data-driven methods has been hindered by a lack of standard typology and scarcity of adequate data. In this study, we leverage the clinical articulation of depression to build a typology for social media texts for detecting the severity of depression. It emulates the standard clinical assessment procedure Diagnostic and Statistical Manual of Mental Disorders (DSM-5) and Patient Health Questionnaire (PHQ-9) to encompass subtle indications of depressive disorders from tweets. Along with the typology, we present a new dataset of 40191 tweets labeled by expert annotators. Each tweet is labeled as 'non-depressed' or 'depressed'. Moreover, three severity levels are considered for 'depressed' tweets: (1) mild, (2) moderate, and (3) severe. An associated confidence score is provided with each label to validate the quality of annotation. We examine the quality of the dataset via representing summary statistics while setting strong baseline results using attention-based models like BERT and DistilBERT. Finally, we extensively address the limitations of the study to provide directions for further research.

Governments have to supervise and inspect social economy enterprises (SEEs). However, inspecting all SEEs is not possible due to the large number of SEEs and the low number of inspectors in general. We proposed a prediction model based on a machine learning approach. The method was trained with the random forest algorithm with historical data provided by each SEE. Three consecutive periods of data were concatenated. The proposed method uses these periods as input data and predicts the risk of each SEE in the fourth period. The model achieved 76\% overall accuracy. In addition, it obtained good accuracy in predicting the high risk of a SEE. We found that the legal nature and the variation of the past-due portfolio are good predictors of the future risk of a SEE. Thus, the risk of a SEE in a future period can be predicted by a supervised machine learning method. Predicting the high risk of a SEE improves the daily work of each inspector by focusing only on high-risk SEEs.

A major open problem in neural machine translation (NMT) is the translation of idiomatic expressions, such as "under the weather". The meaning of these expressions is not composed by the meaning of their constituent words, and NMT models tend to translate them literally (i.e., word-by-word), which leads to confusing and nonsensical translations. Research on idioms in NMT is limited and obstructed by the absence of automatic methods for quantifying these errors. In this work, first, we propose a novel metric for automatically measuring the frequency of literal translation errors without human involvement. Equipped with this metric, we present controlled translation experiments with models trained in different conditions (with/without the test-set idioms) and across a wide range of (global and targeted) metrics and test sets. We explore the role of monolingual pretraining and find that it yields substantial targeted improvements, even without observing any translation examples of the test-set idioms. In our analysis, we probe the role of idiom context. We find that the randomly initialized models are more local or "myopic" as they are relatively unaffected by variations of the idiom context, unlike the pretrained ones.

We present a very simple algorithm for attention that requires $O(1)$ memory with respect to sequence length and an extension to self-attention that requires $O(\log n)$ memory. This is in contrast with the frequently stated belief that self-attention requires $O(n^2)$ memory. While the time complexity is still $O(n^2)$, device memory rather than compute capability is often the limiting factor on modern accelerators. Thus, reducing the memory requirements of attention allows processing of longer sequences than might otherwise be feasible. We provide a practical implementation for accelerators that requires $O(\sqrt{n})$ memory, is numerically stable, and is within a few percent of the runtime of the standard implementation of attention. We also demonstrate how to differentiate the function while remaining memory-efficient. For sequence length 16384, the memory overhead of self-attention is reduced by 59X for inference and by 32X for differentiation.

Federated distillation (FD) paradigm has been recently proposed as a promising alternative to federated learning (FL) especially in wireless sensor networks with limited communication resources. However, all state-of-the art FD algorithms are designed for only classification tasks and less attention has been given to regression tasks. In this work, we propose an FD framework that properly operates on regression learning problems. Afterwards, we present a use-case implementation by proposing an indoor localization system that shows a good trade-off communication load vs. accuracy compared to federated learning (FL) based indoor localization. With our proposed framework, we reduce the number of transmitted bits by up to 98%. Moreover, we show that the proposed framework is much more scalable than FL, thus more likely to cope with the expansion of wireless networks.

Research in massively multilingual image captioning has been severely hampered by a lack of high-quality evaluation datasets. In this paper we present the Crossmodal-3600 dataset (XM3600 in short), a geographically diverse set of 3600 images annotated with human-generated reference captions in 36 languages. The images were selected from across the world, covering regions where the 36 languages are spoken, and annotated with captions that achieve consistency in terms of style across all languages, while avoiding annotation artifacts due to direct translation. We apply this benchmark to model selection for massively multilingual image captioning models, and show superior correlation results with human evaluations when using XM3600 as golden references for automatic metrics.

In this paper, we introduce a novel neural architecture that can judge the result of extracted structured information from documents Measuring the confidence of AI models is critical for safely deploying provided by the information extracting neural networks AI in real-world industrial systems. One important application (hereafter referred to as the IE Networks). Our architecture is hybrid, of confidence measurement is information extraction from scanned consisting of two models, which are a Multi-modal Conformal documents. However, there exists no solution to provide reliable Predictor (MCP) and an Variational Cluster-oriented Anomaly Detector confidence score for current state-of-the-art deep-learning-based (VCAD). The former aims to combine the neural signals from information extractors. In this paper, we propose a complete and 3 main stages of information extraction processes including textbox novel architecture to measure confidence of current deep learning localization, OCR, and key-value recognition to predict the models in document information extraction task. Our architecture confidence level for each extracted key-value output. The later computes consists of a Multi-modal Conformal Predictor and a Variational anomaly scores for the raw input document image, providing Cluster-oriented Anomaly Detector, trained to faithfully estimate the MCP with additional features to produce better confidence estimation.

Fast and precise Lipschitz constant estimation of neural networks is an important task for deep learning. Researchers have recently found an intrinsic trade-off between the accuracy and smoothness of neural networks, so training a network with a loose Lipschitz constant estimation imposes a strong regularization and can hurt the model accuracy significantly. In this work, we provide a unified theoretical framework, a quantitative geometric approach, to address the Lipschitz constant estimation. By adopting this framework, we can immediately obtain several theoretical results, including the computational hardness of Lipschitz constant estimation and its approximability. Furthermore, the quantitative geometric perspective can also provide some insights into recent empirical observations that techniques for one norm do not usually transfer to another one. We also implement the algorithms induced from this quantitative geometric approach in a tool GeoLIP. These algorithms are based on semidefinite programming (SDP). Our empirical evaluation demonstrates that GeoLIP is more scalable and precise than existing tools on Lipschitz constant estimation for $\ell_\infty$-perturbations. Furthermore, we also show its intricate relations with other recent SDP-based techniques, both theoretically and empirically. We believe that this unified quantitative geometric perspective can bring new insights and theoretical tools to the investigation of neural-network smoothness and robustness.

AI in manufacturing is the intelligence of machines to perform humanlike tasks -- responding to events internally and externally, indeed anticipating events autonomously. The machines can determine a tool wearing out or something unanticipated -- perhaps indeed something anticipated to be -- and they can respond and work around the problem. Manufacturers and artificial intelligence service provider are constantly working to identify patterns and work on problems, knowing that indeed the lowest advancements have big implications. They've always been settlers in making smarter use of robotization, so it seems logical that the automated learning that characterizes AI would discover a natural affinity with manufacturing. Yet indeed with that clear coordination, manufacturers have frequently faced challenges to AI adoption.