Objective: Machine learning techniques have been used extensively for 12-lead electrocardiogram (ECG) analysis. For physiological time series, deep learning (DL) superiority to feature engineering (FE) approaches based on domain knowledge is still an open question. Moreover, it remains unclear whether combining DL with FE may improve performance. Methods: We considered three tasks intending to address these research gaps: cardiac arrhythmia diagnosis (multiclass-multilabel classification), atrial fibrillation risk prediction (binary classification), and age estimation (regression). We used an overall dataset of 2.3M 12-lead ECG recordings to train the following models for each task: i) a random forest taking the FE as input was trained as a classical machine learning approach; ii) an end-to-end DL model; and iii) a merged model of FE+DL. Results: FE yielded comparable results to DL while necessitating significantly less data for the two classification tasks and it was outperformed by DL for the regression task. For all tasks, merging FE with DL did not improve performance over DL alone. Conclusion: We found that for traditional 12-lead ECG based diagnosis tasks DL did not yield a meaningful improvement over FE, while it improved significantly the nontraditional regression task. We also found that combining FE with DL did not improve over DL alone which suggests that the FE were redundant with the features learned by DL. Significance: Our findings provides important recommendations on what machine learning strategy and data regime to chose with respect to the task at hand for the development of new machine learning models based on the 12-lead ECG.

We are a technology company working to transform credit and banking in Latin America starting with point of sale finance. We aim to build fair, simple and affordable financial services that empower our clients, treating them with dignity, and building financial freedom. We launched in February 2019 and have already served thousands of clients and disbursed millions of dollars. We operate as a full stack startup. We have built our core systems and processes from scratch, and we believe our technology and analytics platform will drive our progress into the years to come.

A brand new partnership goals to take an enormous chew out of the scourge of single-use plastic. With the assistance of some robots and economies of scale, it is a constructive step in an issue that appears wholly intractable. Compostable packaging firm Zume, which has innovated the manufacture of excessive output molded fiber packaging utilizing robots, is becoming a member of forces with sustainable packaging firm Transcend Packaging. By becoming a member of forces, the businesses try to rally an efficient base of capabilities and distribution to tackle the unfathomable may of the plastics business. At the moment, single-use plastic is a $320B business.

Recent works have shown that attaching prompts to the input is effective at conditioning Language Models (LM) to perform specific tasks. However, prompts are always included in the input text during inference, thus incurring substantial computational and memory overhead. Also, there is currently no straightforward method of utilizing prompts that are longer than the maximum input length of the LMs without incurring additional costs during inference. We propose Prompt Injection (PI), a novel formulation of injecting the prompt into the parameters of an LM to be an efficient alternative to attaching fixed prompts to the input. We show that in scenarios with long fixed prompts, PI can be up to 280 times more efficient in terms of total FLOPs than previous approaches. We further explore methodologies for PI and show promising results in persona-dependent conversation, semantic parsing, and zero-shot learning with task instructions. Through these explorations, we show that PI can be a promising direction for conditioning language models, especially in scenarios with long and fixed prompts.

The classical hinge-loss support vector machines (SVMs) model is sensitive to outlier observations due to the unboundedness of its loss function. To circumvent this issue, recent studies have focused on non-convex loss functions, such as the hard-margin loss, which associates a constant penalty to any misclassified or within-margin sample. Applying this loss function yields much-needed robustness for critical applications but it also leads to an NP-hard model that makes training difficult, since current exact optimization algorithms show limited scalability, whereas heuristics are not able to find high-quality solutions consistently. Against this background, we propose new integer programming strategies that significantly improve our ability to train the hard-margin SVM model to global optimality. We introduce an iterative sampling and decomposition approach, in which smaller subproblems are used to separate combinatorial Benders' cuts. Those cuts, used within a branch-and-cut algorithm, permit to converge much more quickly towards a global optimum. Through extensive numerical analyses on classical benchmark data sets, our solution algorithm solves, for the first time, 117 new data sets to optimality and achieves a reduction of 50% in the average optimality gap for the hardest datasets of the benchmark.

Deep neural networks have been well-known for their superb handling of various machine learning and artificial intelligence tasks. However, due to their over-parameterized black-box nature, it is often difficult to understand the prediction results of deep models. In recent years, many interpretation tools have been proposed to explain or reveal how deep models make decisions. In this paper, we review this line of research and try to make a comprehensive survey. Specifically, we first introduce and clarify two basic concepts -- interpretations and interpretability -- that people usually get confused about. To address the research efforts in interpretations, we elaborate the designs of a number of interpretation algorithms, from different perspectives, by proposing a new taxonomy. Then, to understand the interpretation results, we also survey the performance metrics for evaluating interpretation algorithms. Further, we summarize the current works in evaluating models' interpretability using "trustworthy" interpretation algorithms. Finally, we review and discuss the connections between deep models' interpretations and other factors, such as adversarial robustness and learning from interpretations, and we introduce several open-source libraries for interpretation algorithms and evaluation approaches.

In this paper, we report a hierarchical deep learning model for classification of complex human activities using motion sensors. In contrast to traditional Human Activity Recognition (HAR) models used for event-based activity recognition, such as step counting, fall detection, and gesture identification, this new deep learning model, which we refer to as CHARM (Complex Human Activity Recognition Model), is aimed for recognition of high-level human activities that are composed of multiple different low-level activities in a non-deterministic sequence, such as meal preparation, house chores, and daily routines. CHARM not only quantitatively outperforms state-of-the-art supervised learning approaches for high-level activity recognition in terms of average accuracy and F1 scores, but also automatically learns to recognize low-level activities, such as manipulation gestures and locomotion modes, without any explicit labels for such activities. This opens new avenues for Human-Machine Interaction (HMI) modalities using wearable sensors, where the user can choose to associate an automated task with a high-level activity, such as controlling home automation (e.g., robotic vacuum cleaners, lights, and thermostats) or presenting contextually relevant information at the right time (e.g., reminders, status updates, and weather/news reports). In addition, the ability to learn low-level user activities when trained using only high-level activity labels may pave the way to semi-supervised learning of HAR tasks that are inherently difficult to label.

Present air traffic complexity metrics are defined considering the interests of different management layers of ATM. These layers have different objectives which in practice compete to maximize their own goals, which leads to fragmented decision making. This fragmentation together with competing KPAs requires transparent and neutral air traffic information to pave the way for an explainable set of actions. In this paper, we introduce the concept of single aircraft complexity, to determine the contribution of each aircraft to the overall complexity of air traffic. Furthermore, we describe a methodology extending this concept to define complex communities, which are groups of interdependent aircraft that contribute the majority of the complexity in a certain airspace. In order to showcase the methodology, a tool that visualizes different outputs of the algorithm is developed. Through use-cases based on synthetic and real historical traffic, we first show that the algorithm can serve to formalize controller decisions as well as guide controllers to better decisions. Further, we investigate how the provided information can be used to increase transparency of the decision makers towards different airspace users, which serves also to increase fairness and equity. Lastly, a sensitivity analysis is conducted in order to systematically analyse how each input affects the methodology.

In the recent years, Distributed Denial of Service (DDoS) attacks on Internet of Things (IoT) devices have become one of the prime concerns to Internet users around the world. One of the sources of the attacks on IoT ecosystems are botnets. Intruders force IoT devices to become unavailable for its legitimate users by sending large number of messages within a short interval. This study proposes NFDLM, a lightweight and optimised Artificial Neural Network (ANN) based Distributed Denial of Services (DDoS) attack detection framework with mutual correlation as feature selection method which produces a superior result when compared with Long Short Term Memory (LSTM) and simple ANN. Overall, the detection performance achieves approximately 99\% accuracy for the detection of attacks from botnets. In this work, we have designed and compared four different models where two are based on ANN and the other two are based on LSTM to detect the attack types of DDoS.

Flapping wings are a bio-inspired method to produce lift and thrust in aerial robots, leading to quiet and efficient motion. The advantages of this technology are safety and maneuverability, and physical interaction with the environment, humans, and animals. However, to enable substantial applications, these robots must perch and land. Despite recent progress in the perching field, flapping-wing vehicles, or ornithopters, are to this day unable to stop their flight on a branch. In this paper, we present a novel method that defines a process to reliably and autonomously land an ornithopter on a branch. This method describes the joint operation of a flapping-flight controller, a close-range correction system and a passive claw appendage. Flight is handled by a triple pitch-yaw-altitude controller and integrated body electronics, permitting perching at 3 m/s. The close-range correction system, with fast optical branch sensing compensates for position misalignment when landing. This is complemented by a passive bistable claw design can lock and hold 2 Nm of torque, grasp within 25 ms and can re-open thanks to an integrated tendon actuation. The perching method is supplemented by a four-step experimental development process which optimizes for a successful design. We validate this method with a 700 g ornithopter and demonstrate the first autonomous perching flight of a flapping-wing robot on a branch, a result replicated with a second robot. This work paves the way towards the application of flapping-wing robots for long-range missions, bird observation, manipulation, and outdoor flight.