Predictive models can fail to generalize from training to deployment environments because of dataset shift, posing a threat to model reliability and the safety of downstream decisions made in practice. Instead of using samples from the target distribution to reactively correct dataset shift, we use graphical knowledge of the causal mechanisms relating variables in a prediction problem to proactively remove relationships that do not generalize across environments, even when these relationships may depend on unobserved variables (violations of the "no unobserved confounders" assumption). To accomplish this, we identify variables with unstable paths of statistical influence and remove them from the model. We also augment the causal graph with latent counterfactual variables that isolate unstable paths of statistical influence, allowing us to retain stable paths that would otherwise be removed. Our experiments demonstrate that models that remove vulnerable variables and use estimates of the latent variables transfer better, often outperforming in the target domain despite some accuracy loss in the training domain.

Transportation modes prediction is a fundamental task for decision making in smart cities and traffic management systems. Traffic policies designed based on trajectory mining can save money and time for authorities and the public. It may reduce the fuel consumption and commute time and moreover, may provide more pleasant moments for residents and tourists. Since the number of features that may be used to predict a user transportation mode can be substantial, finding a subset of features that maximizes a performance measure is worth investigating. In this work, we explore wrapper and information retrieval methods to find the best subset of trajectory features. After finding the best classifier and the best feature subset, our results were compared with two related papers that applied deep learning methods and the results showed that our framework achieved better performance. Furthermore, two types of cross-validation approaches were investigated, and the performance results show that the random cross-validation method provides optimistic results.

Because of their effectiveness in broad practical applications, LSTM networks have received a wealth of coverage in scientific journals, technical blogs, and implementation guides. However, in most articles, the inference formulas for the LSTM network and its parent, RNN, are stated axiomatically, while the training formulas are omitted altogether. In addition, the technique of "unrolling" an RNN is routinely presented without justification throughout the literature. The goal of this paper is to explain the essential RNN and LSTM fundamentals in a single document. Drawing from concepts in signal processing, we formally derive the canonical RNN formulation from differential equations. We then propose and prove a precise statement, which yields the RNN unrolling technique. We also review the difficulties with training the standard RNN and address them by transforming the RNN into the "Vanilla LSTM" network through a series of logical arguments. We provide all equations pertaining to the LSTM system together with detailed descriptions of its constituent entities. Albeit unconventional, our choice of notation and the method for presenting the LSTM system emphasizes ease of understanding. As part of the analysis, we identify new opportunities to enrich the LSTM system and incorporate these extensions into the Vanilla LSTM network, producing the most general LSTM variant to date. The target reader has already been exposed to RNNs and LSTM networks through numerous available resources and is open to an alternative pedagogical approach. A Machine Learning practitioner seeking guidance for implementing our new augmented LSTM model in software for experimentation and research will find the insights and derivations in this tutorial valuable as well.

We argue that logical semantics might have faltered due to its failure in distinguishing between two fundamentally very different types of concepts: ontological concepts, that should be types in a strongly-typed ontology, and logical concepts, that are predicates corresponding to properties of and relations between objects of various ontological types. We will then show that accounting for these differences amounts to the integration of lexical and compositional semantics in one coherent framework, and to an embedding in our logical semantics of a strongly-typed ontology that reflects our commonsense view of the world and the way we talk about it in ordinary language. We will show that in such a framework a number of challenges in natural language semantics can be adequately and systematically treated.

We study the problem of finite ontology mediated query answering (FOMQA), the variant of OMQA where the represented world is assumed to be finite, and thus only finite models of the ontology are considered. We adopt the most typical setting with unions of conjunctive queries and ontologies expressed in description logics (DLs). The study of FOMQA is relevant in settings that are not finitely controllable. This is the case not only for DLs without the finite model property, but also for those allowing transitive role declarations. When transitive roles are allowed, evaluating queries is challenging: FOMQA is undecidable for SHOIF and only known to be decidable for the Horn fragment of ALCIF. We show decidability of FOMQA for three proper fragments of SOIF: SOI, SOF, and SIF. Our approach is to characterise models relevant for deciding finite query entailment. Relying on a certain regularity of these models, we develop automata-based decision procedures with optimal complexity bounds.

Working memory is an essential component of reasoning -- the capacity to answer a new question by manipulating acquired knowledge. Current memory-augmented neural networks offer a differentiable method to realize limited reasoning with support of a working memory module. Memory modules are often implemented as a set of memory slots without explicit relational exchange of content. This does not naturally match multi-relational domains in which data is structured. We design a new model dubbed Relational Dynamic Memory Network (RDMN) to fill this gap. The memory can have a single or multiple components, each of which realizes a multi-relational graph of memory slots. The memory is dynamically updated in the reasoning process controlled by the central controller. We evaluate the capability of RDMN on several important application domains: software vulnerability, molecular bioactivity and chemical reaction. Results demonstrate the efficacy of the proposed model.

Algorithm selection and hyperparameter tuning remain two of the most challenging tasks in machine learning. The number of machine learning applications is growing much faster than the number of machine learning experts, hence we see an increasing demand for efficient automation of learning processes. Here, we introduce OBOE, an algorithm for time-constrained model selection and hyperparameter tuning. Taking advantage of similarity between datasets, OBOE finds promising algorithm and hyperparameter configurations through collaborative filtering. Our system explores these models under time constraints, so that rapid initializations can be provided to warm-start more fine-grained optimization methods. One novel aspect of our approach is a new heuristic for active learning in time-constrained matrix completion based on optimal experiment design. Our experiments demonstrate that OBOE delivers state-of-the-art performance faster than competing approaches on a test bed of supervised learning problems.

We present a regression technique for data driven problems based on polynomial chaos expansion (PCE). PCE is a popular technique in the field of uncertainty quantification (UQ), where it is typically used to replace a runnable but expensive computational model subject to random inputs with an inexpensive-to-evaluate polynomial function. The metamodel obtained enables a reliable estimation of the statistics of the output, provided that a suitable probabilistic model of the input is available. In classical machine learning (ML) regression settings, however, the system is only known through observations of its inputs and output, and the interest lies in obtaining accurate pointwise predictions of the latter. Here, we show that a PCE metamodel purely trained on data can yield pointwise predictions whose accuracy is comparable to that of other ML regression models, such as neural networks and support vector machines. The comparisons are performed on benchmark datasets available from the literature. The methodology also enables the quantification of the output uncertainties and is robust to noise. Furthermore, it enjoys additional desirable properties, such as good performance for small training sets and simplicity of construction, with only little parameter tuning required. In the presence of statistically dependent inputs, we investigate two ways to build the PCE, and show through simulations that one approach is superior to the other in the stated settings.

It was going to be the factory of the future. Dubbed the "Alien Dreadnought," Tesla's new manufacturing facility in Fremont, California, was designed to be fully automated -- no humans need apply. If all went well, AI-powered robots would enable the company to achieve a weekly production of 5,000 Model 3 electric cars to keep up with burgeoning demand. But Tesla fell far short of that mark, manufacturing just 2,000 vehicles a week. The problem, as the company painfully discovered, was that full automation wasn't everything it was cracked up to be.

This sounds like easily-dismissible bunkum, but as traditional attempts to explain consciousness continue to fail, the "panpsychist" view is increasingly being taken seriously by credible philosophers, neuroscientists, and physicists, including figures such as neuroscientist Christof Koch and physicist Roger Penrose. "Why should we think common sense is a good guide to what the universe is like?" says Philip Goff, a philosophy professor at Central European University in Budapest, Hungary. "Einstein tells us weird things about the nature of time that counters common sense; quantum mechanics runs counter to common sense. David Chalmers, a philosophy of mind professor at New York University, laid out the "hard problem of consciousness" in 1995, demonstrating that there was still no answer to the question of what causes consciousness. Traditionally, two dominant perspectives, materialism and dualism, have provided a framework for solving this problem. The materialist viewpoint states that consciousness is derived entirely from physical matter. It's unclear, though, exactly how this could work. "It's very hard to get consciousness out of non-consciousness," says Chalmers. It can explain biology, but there's a gap: Consciousness."