We describe the concept of logical scaffolds, which can be used to improve the quality of software that relies on AI components. We explain how some of the existing ideas on runtime monitors for perception systems can be seen as a specific instance of logical scaffolds. Furthermore, we describe how logical scaffolds may be useful for improving AI programs beyond perception systems, to include general prediction systems and agent behavior models. Keywords: AI · Autonomous systems · Formal methods. 1 Introduction Recent progress in AI has led to possible deployment in a wide variety of important domains. This includes safety-critical cyberphysical systems such as automobiles [1] and airplanes [7], but also decision making systems in diverse domains including legal [15] and military applications [3].

Event series, such as the Wimbledon Championships and the US presidential elections, represent important happenings in key societal areas including sports, culture and politics. However, semantic reference sources, such as Wikidata, DBpedia and EventKG knowledge graphs, provide only an incomplete event series representation. In this paper we target the problem of event series completion in a knowledge graph. We address two tasks: 1) prediction of sub-event relations, and 2) inference of real-world events that happened as a part of event series and are missing in the knowledge graph. To address these problems, our proposed supervised HapPenIng approach leverages structural features of event series. HapPenIng does not require any external knowledge - the characteristics making it unique in the context of event inference. Our experimental evaluation demonstrates that HapPenIng outperforms the baselines by 44 and 52 percentage points in terms of precision for the sub-event prediction and the inference tasks, correspondingly. 1 Introduction Event series, such as sports tournaments, music festivals and political elections are sequences of recurring events. Prominent examples include the Wimbledon Championships, the Summer Olympic Games, the United States presidential elections and the International Semantic Web Conference. The provision of reliable reference sources for event series is of crucial importance for many real-world applications, for example in the context of Digital Humanities and Web Science research [7, 9, 25], as well as media analytics and digital journalism [15, 23].

In this paper we consider the basic version of Reinforcement Learning (RL) that involves computing optimal data driven (adaptive) policies for Markovian decision process with unknown transition probabilities. We provide a brief survey of the state of the art of the area and we compare the performance of the classic UCB policy of \cc{bkmdp97} with a new policy developed herein which we call MDP-Deterministic Minimum Empirical Divergence (MDP-DMED), and a method based on Posterior sampling (MDP-PS).

Machine education is an emerging research field that focuses on the problem which is inverse to machine learning. To date, the literature on educating machines is still in its infancy. A fairly low number of methodology and method papers are scattered throughout various formal and informal publication avenues, mainly because the field is not yet well coalesced (with no well established discussion forums or investigation pathways), but also due to the breadth of its potential ramifications and research directions. In this study we bring together the existing literature and organise the discussion into a small number of research directions (out of many) which are to date sufficiently explored to form a minimal critical mass that can push the machine education concept further towards a standalone research field status.

Multi-view clustering is an important yet challenging task due to the difficulty of integrating the information from multiple representations. Most existing multi-view clustering methods explore the heterogeneous information in the space where the data points lie. Such common practice may cause significant information loss because of unavoidable noise or inconsistency among views. Since different views admit the same cluster structure, the natural space should be all partitions. Orthogonal to existing techniques, in this paper, we propose to leverage the multi-view information by fusing partitions. Specifically, we align each partition to form a consensus cluster indicator matrix through a distinct rotation matrix. Moreover, a weight is assigned for each view to account for the clustering capacity differences of views. Finally, the basic partitions, weights, and consensus clustering are jointly learned in a unified framework. We demonstrate the effectiveness of our approach on several real datasets, where significant improvement is found over other state-of-the-art multi-view clustering methods.

Incorporating high-level knowledge is an effective way to expedite reinforcement learning (RL), especially for complex tasks with sparse rewards. We investigate an RL problem where the high-level knowledge is in the form of reward machines, i.e., a type of Mealy machine that encodes the reward functions. We focus on a setting in which this knowledge is a priori not available to the learning agent. We develop an iterative algorithm that performs joint inference of reward machines and policies for RL (more specifically, q-learning). In each iteration, the algorithm maintains a hypothesis reward machine and a sample of RL episodes. It derives q-functions from the current hypothesis reward machine, and performs RL to update the q-functions. While performing RL, the algorithm updates the sample by adding RL episodes along which the obtained rewards are inconsistent with the rewards based on the current hypothesis reward machine. In the next iteration, the algorithm infers a new hypothesis reward machine from the updated sample. Based on an equivalence relationship we defined between states of reward machines, we transfer the q-functions between the hypothesis reward machines in consecutive iterations. We prove that the proposed algorithm converges almost surely to an optimal policy in the limit if a minimal reward machine can be inferred and the maximal length of each RL episode is sufficiently long. The experiments show that learning high-level knowledge in the form of reward machines can lead to fast convergence to optimal policies in RL, while standard RL methods such as q-learning and hierarchical RL methods fail to converge to optimal policies after a substantial number of training steps in many tasks.

We plot the judge's probability of the target answer given that sentence against how often humans also select that target answer given that same sentence. Humans tend to find a sentence to be strong evidence for an answer when the judge model finds it to be strong evidence. Strong evidence to a model tends to be strong evidence to humans as shown in Figure 7. Combined with the previous result, we can see that learned agents are more accurate at predicting sentences that humans find to be strong evidence. F Model Evaluation of Evidence on DREAM Figure 8 shows how convincing various judge models find each evidence agent. Our findings on DREAM are similar to those from RACE in §4.2. Figure 8: On DREAM, how often each judge selects an agent's answer when given a single agent-chosen sentence. The black line divides learned agents (right) and search agents (left), with human evidence selection in the leftmost column. All agents find evidence that convinces judge models more often than a no-evidence baseline (33%). Learned agents predicting p ( i) or p ( i) find the most broadly convincing evidence.

Parameter-transfer is a well-known and versatile approach for meta-learning, with applications including few-shot learning, federated learning, and reinforcement learning. However, parameter-transfer algorithms often require sharing models that have been trained on the samples from specific tasks, thus leaving the task-owners susceptible to breaches of privacy. We conduct the first formal study of privacy in this setting and formalize the notion of task-global differential privacy as a practical relaxation of more commonly studied threat models. We then propose a new differentially private algorithm for gradient-based parameter transfer that not only satisfies this privacy requirement but also retains provable transfer learning guarantees in convex settings. Empirically, we apply our analysis to the problem of federated learning with personalization and show that allowing the relaxation to task-global privacy from the more commonly studied notion of local privacy leads to dramatically increased performance in recurrent neural language modeling.

Neural entity linking models are very powerful, but run the risk of overfitting to the domain they are trained in. For this problem, a domain can be narrowly construed as a particular distribution of entities, as models can even overfit by memorizing properties of specific frequent entities in a dataset. We tackle the problem of building robust entity linking models that generalize effectively and do not rely on labeled entity linking data with a specific entity distribution. Rather than predicting entities directly, our approach models fine-grained entity properties, which can help disambiguate between even closely related entities. We derive a large inventory of types (tens of thousands) from Wikipedia categories, and use hy-perlinked mentions in Wikipedia to distantly label data and train an entity typing model. At test time, we classify a mention with this typing model and use soft type predictions to link the mention to the most similar candidate entity. We evaluate our entity linking system on the CoNLL-Y AGO (Hoffart et al., 2011) dataset and show that our approach outperforms prior domain-independent entity linking systems. We also test our approach in a harder setting derived from the WikilinksNED dataset (Eshel et al., 2017) where all the mention-entity pairs are unseen during test time. Results indicate that our approach generalizes better than a state-of-the-art neural model on the dataset. 1 Introduction Historically, systems for entity linking to Wikipedia relied on heuristics such as anchor text distributions (Cucerzan, 2007; Milne and Witten, 2008), tf-idf (Ratinov et al., 2011), and Wikipedia relatedness of nearby entities (Hoffart et al., 2011). These systems have few parameters, making them relatively flexible across domains. More recent systems have typically been parameter-rich neural network models (Sun et al., 2015; Y amada et al., 2016; Francis-Landau et al., 2016; Eshel et al., 2017).

Conversion of raw data into insights and knowledge requires substantial amounts of effort from data scientists. Despite breathtaking advances in Machine Learning (ML) and Artificial Intelligence (AI), data scientists still spend the majority of their effort in understanding and then preparing the raw data for ML/AI. The effort is often manual and ad hoc, and requires some level of domain knowledge. The complexity of the effort increases dramatically when data diversity, both in form and context, increases. In this paper, we introduce our solution, Augmented Data Science (ADS), towards addressing this "human bottleneck" in creating value from diverse datasets. ADS is a data-driven approach and relies on statistics and ML to extract insights from any data set in a domain-agnostic way to facilitate the data science process. Key features of ADS are the replacement of rudimentary data exploration and processing steps with automation and the augmentation of data scientist judgment with automatically-generated insights. We present building blocks of our end-to-end solution and provide a case study to exemplify its capabilities.