Annotating the driving scenario trajectories based only on explicit rules (i.e., knowledge-based methods) can be subject to errors, such as false positive/negative classification of scenarios that lie on the border of two scenario classes, missing unknown scenario classes, and also anomalies. On the other side, verifying the labels by the annotators is not cost-efficient. For this purpose, active learning (AL) could potentially improve the annotation procedure by inclusion of an annotator/expert in an efficient way. In this study, we develop an active learning framework to annotate driving trajectory time-series data. At the first step, we compute an embedding of the time-series trajectories into a latent space in order to extract the temporal nature. For this purpose, we study three different latent space representations: multivariate Time Series t-Distributed Stochastic Neighbor Embedding (mTSNE), Recurrent Auto-Encoder (RAE) and Variational Recurrent Auto-Encoder (VRAE). We then apply different active learning paradigms with different classification models to the embedded data. In particular, we study the two classifiers Neural Network (NN) and Support Vector Machines (SVM), with three active learning query strategies (i.e., entropy, margin and random). In the following, we explore the possibilities of the framework to discover unknown classes and demonstrate how it can be used to identify the out-of-class trajectories.

The absence of a signal of new particles at the Large Hadron Collider (LHC) may suggest that new physics is realized in a scenario that is hard to detect due to the absence or very large mass of new colored particles. Hence, this study focuses on setups with dominant electroweak production of color-neutral new particles and multi-lepton signals from their decays. The conventional approach to searches for new physics, also known as "cut-and-count analysis", is to apply a set of constraints on different kinematic variables (called "cuts" or "selection") that improve the signalto-background ratio. However, the scenarios we consider can be challenging for this standard approach due to the small production cross section and the similarity of signal and background features. For such problems, machine learning (ML) offers a promising alternative [1-6]. We investigate how much ML can increase the discovery reach, and whether machine learning models can be trained in such a way that they work in a large region of parameter space and not just for a single point. This is an important issue, in particular in new physics scenarios with many free parameters, as signal kinematics vary from point to point. As a concrete example, we consider a supersymmetry (SUSY) scenario with a gravitino lightest supersymmetric particle (LSP) whose mass is in the GeV range.

We consider the problem of estimating latent positions in a one-dimensional torus from pairwise affinities. The observed affinity between a pair of items is modeled as a noisy observation of a function $f(x^*_{i},x^*_{j})$ of the latent positions $x^*_{i},x^*_{j}$ of the two items on the torus. The affinity function $f$ is unknown, and it is only assumed to fulfill some shape constraints ensuring that $f(x,y)$ is large when the distance between $x$ and $y$ is small, and vice-versa. This non-parametric modeling offers a good flexibility to fit data. We introduce an estimation procedure that provably localizes all the latent positions with a maximum error of the order of $\sqrt{\log(n)/n}$, with high-probability. This rate is proven to be minimax optimal. A computationally efficient variant of the procedure is also analyzed under some more restrictive assumptions. Our general results can be instantiated to the problem of statistical seriation, leading to new bounds for the maximum error in the ordering.

Conditional average treatment effects (CATEs) allow us to understand the effect heterogeneity across a large population of individuals. However, typical CATE learners assume all confounding variables are measured in order for the CATE to be identifiable. Often, this requirement is satisfied by simply collecting many variables, at the expense of increased sample complexity for estimating CATEs. To combat this, we propose an energy-based model (EBM) that learns a low-dimensional representation of the variables by employing a noise contrastive loss function. With our EBM we introduce a preprocessing step that alleviates the dimensionality curse for any existing model and learner developed for estimating CATE. We prove that our EBM keeps the representations partially identifiable up to some universal constant, as well as having universal approximation capability to avoid excessive information loss from model misspecification; these properties combined with our loss function, enable the representations to converge and keep the CATE estimation consistent. Experiments demonstrate the convergence of the representations, as well as show that estimating CATEs on our representations performs better than on the variables or the representations obtained via various benchmark dimensionality reduction methods.

The need of predictive maintenance comes with an increasing number of incidents reported by monitoring systems and equipment/software users. In the front line, on-call engineers (OCEs) have to quickly assess the degree of severity of an incident and decide which service to contact for corrective actions. To automate these decisions, several predictive models have been proposed, but the most efficient models are opaque (say, black box), strongly limiting their adoption. In this paper, we propose an efficient black box model based on 170K incidents reported to our company over the last 7 years and emphasize on the need of automating triage when incidents are massively reported on thousands of servers running our product, an ERP. Recent developments in eXplainable Artificial Intelligence (XAI) help in providing global explanations to the model, but also, and most importantly, with local explanations for each model prediction/outcome. Sadly, providing a human with an explanation for each outcome is not conceivable when dealing with an important number of daily predictions. To address this problem, we propose an original data-mining method rooted in Subgroup Discovery, a pattern mining technique with the natural ability to group objects that share similar explanations of their black box predictions and provide a description for each group. We evaluate this approach and present our preliminary results which give us good hope towards an effective OCE's adoption. We believe that this approach provides a new way to address the problem of model agnostic outcome explanation.

The goal of representation learning is different from the ultimate objective of machine learning such as decision making, it is therefore very difficult to establish clear and direct objectives for training representation learning models. It has been argued that a good representation should disentangle the underlying variation factors, yet how to translate this into training objectives remains unknown. This paper presents an attempt to establish direct training criterions and design principles for developing good representation learning models. We propose that a good representation learning model should be maximally expressive, i.e., capable of distinguishing the maximum number of input configurations. We formally define expressiveness and introduce the maximum expressiveness (MEXS) theorem of a general learning model. We propose to train a model by maximizing its expressiveness while at the same time incorporating general priors such as model smoothness. We present a conscience competitive learning algorithm which encourages the model to reach its MEXS whilst at the same time adheres to model smoothness prior. We also introduce a label consistent training (LCT) technique to boost model smoothness by encouraging it to assign consistent labels to similar samples. We present extensive experimental results to show that our method can indeed design representation learning models capable of developing representations that are as good as or better than state of the art. We also show that our technique is computationally efficient, robust against different parameter settings and can work effectively on a variety of datasets. Code available at https://github.com/qlilx/odgrlm.git

As machine learning models grow more complex and their applications become more high-stakes, tools for explaining model predictions have become increasingly important. This has spurred a flurry of research in model explainability and has given rise to feature attribution methods such as LIME and SHAP. Despite their widespread use, evaluating and comparing different feature attribution methods remains challenging: evaluations ideally require human studies, and empirical evaluation metrics are often data-intensive or computationally prohibitive on real-world datasets. In this work, we address this issue by releasing XAI-Bench: a suite of synthetic datasets along with a library for benchmarking feature attribution algorithms. Unlike real-world datasets, synthetic datasets allow the efficient computation of conditional expected values that are needed to evaluate ground-truth Shapley values and other metrics. The synthetic datasets we release offer a wide variety of parameters that can be configured to simulate real-world data. We demonstrate the power of our library by benchmarking popular explainability techniques across several evaluation metrics and across a variety of settings. The versatility and efficiency of our library will help researchers bring their explainability methods from development to deployment. Our code is available at https://github.com/abacusai/xai-bench.

Becoming a data scientist is considered a prestigious trait. Back in 2012, Harvard Business Review called'data scientist' the sexiest job of the 21st century, and the growing trend of roles in the industry seems to be confirming that statement. To confirm this sexiness is still ongoing, the info from Glassdoor shows being a data scientist is the second-best job in America in 2021. To get such a prestigious job, you have to go through rigorous job interviews. Data science questions asked can be very broad and complex. This is expected, considering the role of a data scientist usually incorporates so many areas.

As battles to contain the COVID-19 pandemic continue, attention is focused on emerging variants of the severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) virus that have been deemed variants of concern because they are resistant to antibodies elicited by infection or vaccination or they increase transmissibility or disease severity. Three papers used functional and structural studies to explore how mutations in the viral spike protein affect its ability to infect host cells and to evade host immunity. Gobeil et al. looked at a variant spike protein involved in transmission between minks and humans, as well as the B1.1.7 (alpha), B.1.351 (beta), and P1 (gamma) spike variants; Cai et al. focused on the alpha and beta variants; and McCallum et al. discuss the properties of the spike protein from the B1.1.427/B.1.429 (epsilon) variant. Together, these papers show a balance among mutations that enhance stability, those that increase binding to the human receptor ACE2, and those that confer resistance to neutralizing antibodies. Science , abi6226, abi9745, abi7994, this issue p. [eabi6226][1] , p. [642][2], p. [648][3] ### INTRODUCTION Variants of severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) have been circulating worldwide since the beginning of the pandemic. Some are termed Variants of Concern (VOC) because they show evidence for increased transmissibility, higher disease severity, resistance to neutralizing antibodies elicited by current vaccines or from previous infection, reduced efficacy of treatments, or failure of diagnostic detection methods. VOCs accumulate mutations in the spike (S) glycoprotein. Some VOCs that arose independently in different geographical locations show identical changes, implying convergent evolution and selective advantages of the acquired variations. A set of three amino acid substitutions in the receptor-binding domain (RBD)—Lys417 → Asn (K417N), Glu484 → Lys (E484K), and Asn501 → Tyr (N501Y)—occurred in the B.1.1.28 and B.1.351 lineages that originated in Brazil and South Africa, respectively. The P.1 lineage that branched off B.1.1.28 harbored a Lys417 → Thr (K417T) substitution while retaining the E484K and N501Y changes. The E484K substitution has attracted attention as a result of its location within the epitope of many potent neutralizing antibodies. The N501Y substitution also occurred in the B.1.1.7 variant that originated in the UK and was implicated in increased receptor binding and higher transmissibility of the variant. The B.1.1.7 variant, in turn, shares the His69/Val70 spike deletion mutation with spike from a variant that was implicated in transmission between humans and minks (ΔFVI). ### RATIONALE Global sequencing initiatives and in vitro neutralization and antibody binding assays have rapidly provided critical and timely information on the VOCs. Here, by combining cryo–electron microscopy (cryo-EM) structural determination with binding assays and computational analyses on the variant spikes, we sought to visualize the impact of the amino acid substitutions on spike conformation to understand how these changes affect their biological function. ### RESULTS We measured angiotensin-converting enzyme 2 (ACE2) receptor and antibody binding for 19 SARS-CoV-2 S ectodomain constructs harboring amino acid changes found in circulating variants. These included a variant involved in interspecies SARS-CoV-2 transmission between humans and minks, as well as several VOCs including the B.1.1.7, B.1.1.28/P.1, and B.1.351 variants. Consistent with published neutralization data, B.1.1.7 showed decreased binding to N-terminal domain (NTD)–directed antibodies, whereas P.1 and B.1.351 showed reduced binding to both NTD- and RBD-directed antibodies. All variants showed increased binding to ACE2, which was mediated by higher propensity for RBD-up states, and affinity-enhancing mutations in the RBD. We observed spike instability in the mink-associated variant, highlighted by the presence of a population in the cryo-EM dataset with missing density for the S1 subunit of one protomer. Modulation of contacts between the SD1 and HR1 regions led to increased RBD-up states of the B.1.1.7 spike, with the protein stability maintained by a balance of stabilizing and destabilizing mutations. A local destabilizing effect of the RBD E484K mutation was implicated in resistance of the B.1.1.28/P.1 and B.1.351 variants to some potent RBD-directed neutralizing antibodies. ### CONCLUSION Our study revealed details of how amino acid substitutions affect spike conformation in circulating SARS-CoV-2 VOCs. We define communication networks that modulate spike allostery and show that the S protein uses different mechanisms to converge upon similar solutions for altering the RBD up/down positioning. ![Figure][4] Cryo-EM structures of SARS-CoV-2 spike ectodomains. Naturally occurring amino acid variations are represented by colored spheres. Spike mutations from a mink-associated (ΔFV) (top left), B.1.1.7 (top right), B.1.351 (bottom right), and a spike with three RBD mutations (bottom left) are shown. Relative proportions of the RBD down and up populations are indicated for each. The three amino acid substitutions in the RBD—K417N/T, E484K, and N501Y—were found in the B.1.1.28 variant and are shared with the P.1 and B.1.351 lineages. Severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) variants with multiple spike mutations enable increased transmission and antibody resistance. We combined cryo–electron microscopy (cryo-EM), binding, and computational analyses to study variant spikes, including one that was involved in transmission between minks and humans, and others that originated and spread in human populations. All variants showed increased angiotensin-converting enzyme 2 (ACE2) receptor binding and increased propensity for receptor binding domain (RBD)–up states. While adaptation to mink resulted in spike destabilization, the B.1.1.7 (UK) spike balanced stabilizing and destabilizing mutations. A local destabilizing effect of the RBD E484K mutation was implicated in resistance of the B.1.1.28/P.1 (Brazil) and B.1.351 (South Africa) variants to neutralizing antibodies. Our studies revealed allosteric effects of mutations and mechanistic differences that drive either interspecies transmission or escape from antibody neutralization. [1]: /lookup/doi/10.1126/science.abi6226 [2]: /lookup/doi/10.1126/science.abi9745 [3]: /lookup/doi/10.1126/science.abi7994 [4]: pending:yes

While the forward and backward modeling of the process-structure-property chain has received a lot of attention from the materials community, fewer efforts have taken into consideration uncertainties. Those arise from a multitude of sources and their quantification and integration in the inversion process are essential in meeting the materials design objectives. The first contribution of this paper is a flexible, fully probabilistic formulation of such optimization problems that accounts for the uncertainty in the process-structure and structure-property linkages and enables the identification of optimal, high-dimensional, process parameters. We employ a probabilistic, data-driven surrogate for the structure-property link which expedites computations and enables handling of non-differential objectives. We couple this with a novel active learning strategy, i.e. a self-supervised collection of data, which significantly improves accuracy while requiring small amounts of training data. We demonstrate its efficacy in optimizing the mechanical and thermal properties of two-phase, random media but envision its applicability encompasses a wide variety of microstructure-sensitive design problems.