Gross domestic product (GDP) is the most widely used indicator in macroeconomics and the main tool for measuring a country's economic ouput. Due to the diversity and complexity of the world economy, a wide range of models have been used, but there are challenges in making decadal GDP forecasts given unexpected changes such as pandemics and wars. Deep learning models are well suited for modeling temporal sequences have been applied for time series forecasting. In this paper, we develop a deep learning framework to forecast the GDP growth rate of the world economy over a decade. We use Penn World Table as the source of our data, taking data from 1980 to 2019, across 13 countries, such as Australia, China, India, the United States and so on. We test multiple deep learning models, LSTM, BD-LSTM, ED-LSTM and CNN, and compared their results with the traditional time series model (ARIMA,VAR). Our results indicate that ED-LSTM is the best performing model. We present a recursive deep learning framework to predict the GDP growth rate in the next ten years. We predict that most countries will experience economic growth slowdown, stagnation or even recession within five years; only China, France and India are predicted to experience stable, or increasing, GDP growth.

In text mining, topic models are a type of probabilistic generative models for inferring latent semantic topics from text corpus. One of the most popular inference approaches to topic models is perhaps collapsed Gibbs sampling (CGS), which typically samples one single topic label for each observed document-word pair. In this paper, we aim at improving the inference of CGS for topic models. We propose to leverage state augmentation technique by maximizing the number of topic samples to infinity, and then develop a new inference approach, called infinite latent state replication (ILR), to generate robust soft topic assignment for each given document-word pair. Experimental results on the publicly available datasets show that ILR outperforms CGS for inference of existing established topic models.

PPO (Proximal Policy Optimization) is a state-of-the-art policy gradient algorithm that has been successfully applied to complex computer games such as Dota 2 and Honor of Kings. In these environments, an agent makes compound actions consisting of multiple sub-actions. PPO uses clipping to restrict policy updates. Although clipping is simple and effective, it is not efficient in its sample use. For compound actions, most PPO implementations consider the joint probability (density) of sub-actions, which means that if the ratio of a sample (state compound-action pair) exceeds the range, the gradient the sample produces is zero. Instead, for each sub-action we calculate the loss separately, which is less prone to clipping during updates thereby making better use of samples. Further, we propose a multi-action mixed loss that combines joint and separate probabilities. We perform experiments in Gym-$\mu$RTS and MuJoCo. Our hybrid model improves performance by more than 50\% in different MuJoCo environments compared to OpenAI's PPO benchmark results. And in Gym-$\mu$RTS, we find the sub-action loss outperforms the standard PPO approach, especially when the clip range is large. Our findings suggest this method can better balance the use-efficiency and quality of samples.

Federated Learning (FL) is a well-established technique for privacy preserving distributed training. Much attention has been given to various aspects of FL training. A growing number of applications that consume FL-trained models, however, increasingly operate under dynamically and unpredictably variable conditions, rendering a single model insufficient. We argue for training a global family of models cost efficiently in a federated fashion. Training them independently for different tradeoff points incurs $O(k)$ cost for any k architectures of interest, however. Straightforward applications of FL techniques to recent weight-shared training approaches is either infeasible or prohibitively expensive. We propose SuperFed - an architectural framework that incurs $O(1)$ cost to co-train a large family of models in a federated fashion by leveraging weight-shared learning. We achieve an order of magnitude cost savings on both communication and computation by proposing two novel training mechanisms: (a) distribution of weight-shared models to federated clients, (b) central aggregation of arbitrarily overlapping weight-shared model parameters. The combination of these mechanisms is shown to reach an order of magnitude (9.43x) reduction in computation and communication cost for training a $5*10^{18}$-sized family of models, compared to independently training as few as $k = 9$ DNNs without any accuracy loss.

Modelling taxonomic and thematic relatedness is important for building AI with comprehensive natural language understanding. The goal of this paper is to learn more about how taxonomic information is structurally encoded in embeddings. To do this, we design a new hypernym-hyponym probing task and perform a comparative probing study of taxonomic and thematic SGNS and GloVe embeddings. Our experiments indicate that both types of embeddings encode some taxonomic information, but the amount, as well as the geometric properties of the encodings, are independently related to both the encoder architecture, as well as the embedding training data. Specifically, we find that only taxonomic embeddings carry taxonomic information in their norm, which is determined by the underlying distribution in the data.

Machine Learning (ML) is being used in multiple disciplines due to its powerful capability to infer relationships within data. In particular, Software Engineering (SE) is one of those disciplines in which ML has been used for multiple tasks, like software categorization, bugs prediction, and testing. In addition to the multiple ML applications, some studies have been conducted to detect and understand possible pitfalls and issues when using ML. However, to the best of our knowledge, only a few studies have focused on presenting ML best practices or guidelines for the application of ML in different domains. In addition, the practices and literature presented in previous literature (i) are domain-specific (e.g., concrete practices in biomechanics), (ii) describe few practices, or (iii) the practices lack rigorous validation and are presented in gray literature. In this paper, we present a study listing 127 ML best practices systematically mining 242 posts of 14 different Stack Exchange (STE) websites and validated by four independent ML experts. The list of practices is presented in a set of categories related to different stages of the implementation process of an ML-enabled system; for each practice, we include explanations and examples. In all the practices, the provided examples focus on SE tasks. We expect this list of practices could help practitioners to understand better the practices and use ML in a more informed way, in particular newcomers to this new area that sits at the intersection of software engineering and machine learning.

Emerging multi-robot systems rely on cooperation between humans and robots, with robots following automatically generated motion plans to service application-level tasks. Given the safety requirements associated with operating in proximity to humans and expensive infrastructure, it is important to understand and mitigate the security vulnerabilities of such systems caused by compromised robots who diverge from their assigned plans. We focus on centralized systems, where a *central entity* (CE) is responsible for determining and transmitting the motion plans to the robots, which report their location as they move following the plan. The CE checks that robots follow their assigned plans by comparing their expected location to the location they self-report. We show that this self-reporting monitoring mechanism is vulnerable to *plan-deviation attacks* where compromised robots don't follow their assigned plans while trying to conceal their movement by mis-reporting their location. We propose a two-pronged mitigation for plan-deviation attacks: (1) an attack detection technique leveraging both the robots' local sensing capabilities to report observations of other robots and *co-observation schedules* generated by the CE, and (2) a prevention technique where the CE issues *horizon-limiting announcements* to the robots, reducing their instantaneous knowledge of forward lookahead steps in the global motion plan. On a large-scale automated warehouse benchmark, we show that our solution enables attack prevention guarantees from a stealthy attacker that has compromised multiple robots.

Conversational information seeking (CIS) is concerned with a sequence of interactions between one or more users and an information system. Interactions in CIS are primarily based on natural language dialogue, while they may include other types of interactions, such as click, touch, and body gestures. This monograph provides a thorough overview of CIS definitions, applications, interactions, interfaces, design, implementation, and evaluation. This monograph views CIS applications as including conversational search, conversational question answering, and conversational recommendation. Our aim is to provide an overview of past research related to CIS, introduce the current state-of-the-art in CIS, highlight the challenges still being faced in the community. and suggest future directions.

In recent years there have been considerable advances in pre-trained language models, where non-English language versions have also been made available. Due to their increasing use, many lightweight versions of these models (with reduced parameters) have also been released to speed up training and inference times. However, versions of these lighter models (e.g., ALBERT, DistilBERT) for languages other than English are still scarce. In this paper we present ALBETO and DistilBETO, which are versions of ALBERT and DistilBERT pre-trained exclusively on Spanish corpora. We train several versions of ALBETO ranging from 5M to 223M parameters and one of DistilBETO with 67M parameters. We evaluate our models in the GLUES benchmark that includes various natural language understanding tasks in Spanish. The results show that our lightweight models achieve competitive results to those of BETO (Spanish-BERT) despite having fewer parameters. More specifically, our larger ALBETO model outperforms all other models on the MLDoc, PAWS-X, XNLI, MLQA, SQAC and XQuAD datasets. However, BETO remains unbeaten for POS and NER. As a further contribution, all models are publicly available to the community for future research.

To promote new scientific discoveries from complex data sets, feature importance inference has been a long-standing statistical problem. Instead of testing for parameters that are only interpretable for specific models, there has been increasing interest in model-agnostic methods, often in the form of feature occlusion or leave-one-covariate-out (LOCO) inference. Existing approaches often make distributional assumptions, which can be difficult to verify in practice, or require model refitting and data splitting, which are computationally intensive and lead to losses in power. In this work, we develop a novel, mostly model-agnostic and distribution-free inference framework for feature importance that is computationally efficient and statistically powerful. Our approach is fast as we avoid model refitting by leveraging a form of random observation and feature subsampling called minipatch ensembles; this approach also improves statistical power by avoiding data splitting. Our framework can be applied on tabular data and with any machine learning algorithm, together with minipatch ensembles, for regression and classification tasks. Despite the dependencies induced by using minipatch ensembles, we show that our approach provides asymptotic coverage for the feature importance score of any model under mild assumptions. Finally, our same procedure can also be leveraged to provide valid confidence intervals for predictions, hence providing fast, simultaneous quantification of the uncertainty of both predictions and feature importance. We validate our intervals on a series of synthetic and real data examples, including non-linear settings, showing that our approach detects the correct important features and exhibits many computational and statistical advantages over existing methods.