Recent works have shown that generative sequence models (e.g., language models) have a tendency to memorize rare or unique sequences in the training data. Since useful models are often trained on sensitive data, to ensure the privacy of the training data it is critical to identify and mitigate such unintended memorization. Federated Learning (FL) has emerged as a novel framework for large-scale distributed learning tasks. However, it differs in many aspects from the well-studied central learning setting where all the data is stored at the central server. In this paper, we initiate a formal study to understand the effect of different components of canonical FL on unintended memorization in trained models, comparing with the central learning setting. Our results show that several differing components of FL play an important role in reducing unintended memorization. Specifically, we observe that the clustering of data according to users---which happens by design in FL---has a significant effect in reducing such memorization, and using the method of Federated Averaging for training causes a further reduction. We also show that training with a strong user-level differential privacy guarantee results in models that exhibit the least amount of unintended memorization.

Recently, a groundswell of research has identified the use of counterfactual explanations as a potentially significant solution to the Explainable AI (XAI) problem. It is argued that (a) technically, these counterfactual cases can be generated by permuting problem-features until a class change is found, (b) psychologically, they are much more causally informative than factual explanations, (c) legally, they are GDPR-compliant. However, there are issues around the finding of good counterfactuals using current techniques (e.g. sparsity and plausibility). We show that many commonly-used datasets appear to have few good counterfactuals for explanation purposes. So, we propose a new case based approach for generating counterfactuals using novel ideas about the counterfactual potential and explanatory coverage of a case-base. The new technique reuses patterns of good counterfactuals, present in a case-base, to generate analogous counterfactuals that can explain new problems and their solutions. Several experiments show how this technique can improve the counterfactual potential and explanatory coverage of case-bases that were previously found wanting.

An ounce of prevention is worth a pound of cure, as the old saying goes. Until recently, that simply meant living a healthy lifestyle, getting regular checkups, and hoping that signs of anything serious were caught early. But today, doctors are using artificial intelligence (AI) and machine learning systems to make preventative care, diagnosis, and treatment more accurate and effective than ever. "Machine learning involves adaptive learning and as such, can identify patterns over time as new data is aggregated and analyzed," explains Melissa Manice, co-founder of healthcare startup Cohero Health. "Therefore, machine learning and AI allows doctors to detect abnormal behaviors and predictive insights with the application of clinical thresholds to machine learning algorithms," she continues.

During the early stages of developing Case-Based Reasoning (CBR) systems the definition of similarity measures is challenging since this task requires to transfer implicit knowledge of domain experts into knowledge representations. While an entire CBR system is very explanatory, the similarity measure determines the ranking but do not necessarily show which features contribute to high (or low) rankings. In this paper we will present our work on opening the knowledge engineering process for similarity modelling. We will present how we transfer implicit knowledge from experts as well as a data-driven approach into case and similarity representations for CBR systems. The work present is a result of interdisciplinary research collaborations between AI and medical researchers developing e-Health applications. During this work, explainability and transparency of the development process is crucial to allow in-depth quality assurance of the by the domain experts.

In this paper, we present RALE-ACL, a communication language for case-based agents in multi-agent systems (MAS) that utilize case-based reasoning (CBR) as the main means of decision making for their agents. RALE-ACL is an accompanying approach of RALE-CBR, a methodology for construction of CBR-based approaches and systems that adds more flexibility to the classic 4R cycle of case-based reasoning. The main goal of RALE-ACL is to establish a much more CBR-compatible alternative to the KQML and FIPA-ACL-based languages, that are currently used in many multi-agent systems, but are too generic and therefore only cumbersomely usable for the specific structure and purposes of case-based agents. This paper is the final part in the trilogy about the RALE methodology.

Researchers seek to identify biological markers which accurately differentiate cancer subtypes and their severity from normal controls. One such biomarker, DNA methylation, has recently become more prevalent in genetic research studies in oncology. This paper proposes to apply these findings in a study of the diagnostic accuracy of DNA methylation signatures for classifying metastasis samples. Very high classification performance measures were obtained from differentially methylated positions and regions, as well as from selected gene signatures. Perfect accuracy was achieved with the top 5 feature-selected genes using three similar cases and the K-nearest neighbor classfier. This work contributes to the path toward the identification of biological signatures for oncology samples using case-based reasoning.

Autonomous agents often have sufficient resources to achieve the goals that are provided to them. However, in dynamic worlds where unexpected problems are bound to occur, an agent may formulate new goals with further resource requirements. Thus, agents should be smart enough to man-age their goals and the limited resources they possess in an effective and flexible manner. We present an approach to the selection and monitoring of goals using resource estimation and goal priorities. To evaluate our approach, we designed an experiment on top of our previous work in a complex mine-clearance domain. The agent in this domain formulates its own goals by retrieving a case to explain uncovered discrepancies and generating goals from the explanation. Finally, we compare the performance of our approach to two alternatives.

Lattice-based motion planning is a hybrid planning method where a plan made up of discrete actions simultaneously is a physically feasible trajectory. The planning takes both discrete and continuous aspects into account, for example action pre-conditions and collision-free action-duration in the configuration space. Safe motion planing rely on well-calibrated safety-margins for collision checking. The trajectory tracking controller must further be able to reliably execute the motions within this safety margin for the execution to be safe. In this work we are concerned with introspective learning and reasoning about controller performance over time. Normal controller execution of the different actions is learned using reliable and uncertainty-aware machine learning techniques. By correcting for execution bias we manage to substantially reduce the safety margin of motion actions. Reasoning takes place to both verify that the learned models stays safe and to improve collision checking effectiveness in the motion planner by the use of more accurate execution predictions with a smaller safety margin. The presented approach allows for explicit awareness of controller performance under normal circumstances, and timely detection of incorrect performance in abnormal circumstances. Evaluation is made on the nonlinear dynamics of a quadcopter in 3D using simulation. Video: https://youtu.be/STmZduvSUMM

Today during its annual IBM Think conference, IBM announced the launch of Watson AIOps, a service that taps AI to automate the real-time detection, diagnosing, and remediation of network anomalies. It also unveiled new offerings targeting the rollout of 5G technologies and the devices on those networks, as well as a coalition of telecommunications partners -- the IBM Telco Network Cloud Ecosystem -- that will work with IBM to deploy edge computing technologies. Watson AIOps marks IBM's foray into the mammoth AIOps market, which is expected to grow from $2.55 billion in 2018 to $11.02 billion by 2023, according to Markets and Markets. That might be a conservative projection in light of the pandemic, which is forcing IT teams to increasingly conduct their work remotely. In lieu of access to infrastructure, tools like Watson AIOps could help prevent major outages, the cost of which a study from Aberdeen pegged at $260,000 per hour. "The COVID-19 crisis and increased demand for remote work capabilities are driving the need for AI automation at an unprecedented rate and pace," said IBM SVP Rob Thomas in a statement.

We introduce kNN-LMs, which extend a pre-trained neural language model (LM) by linearly interpolating it with a k-nearest neighbors (kNN) model. The nearest neighbors are computed according to distance in the pre-trained LM embedding space, and can be drawn from any text collection, including the original LM training data. Applying this augmentation to a strong WIKITEXT-103 LM, with neighbors drawn from the original training set, our kNN-LM achieves a new state-of-the-art perplexity of 15.79 – a 2.9 point improvement with no additional training. We also show that this approach has implications for efficiently scaling up to larger training sets and allows for effective domain adaptation, by simply varying the nearest neighbor datastore, again without further training. Qualitatively, the model is particularly helpful in predicting rare patterns, such as factual knowledge.