From a small number of calls to a given "blackbox" on random input perturbations, we show how to efficiently recover its unknown Jacobian, or estimate the left action of its Jacobian on a given vector.

Weaknesses: The constants in the bounds depend linearly on the dimension, although they depends exponentially on the regularization parameter. If Sinkhorn distance is thought as a proxy of the Wasserstein distance, this seems to be a hidden dependance on the dimension, since the regularization parameter plays the role of an interpolation between MMD and Wasserstein distances, and MMD distances are more blind to the dimension. This is not discussed in the paper. The results also have an exponential dependence on an assumed uniform upper bound on the cost. For the classical quadratic cost, this imply an exponential dependence on the dimension for the case of measures supported on [0,1] d for instance.

From a small number of calls to a given "blackbox" on random input perturbations, we show how to efficiently recover its unknown Jacobian, or estimate the left action of its Jacobian on a given vector. Our methods are based on a novel combination of compressed sensing and graph coloring techniques, and provably exploit structural prior knowledge about the Jacobian such as sparsity and symmetry while being noise robust. We demonstrate efficient backpropagation through noisy blackbox layers in a deep neural net, improved data-efficiency in the task of linearizing the dynamics of a rigid body system, and the generic ability to handle a rich class of input-output dependency structures in Jacobian estimation problems.

ML model design either starts with an interpretable model or a Blackbox and explains it post hoc. Blackbox models are flexible but difficult to explain, while interpretable models are inherently explainable. Yet, interpretable models require extensive ML knowledge and tend to be less flexible and underperforming than their Blackbox variants. This paper aims to blur the distinction between a post hoc explanation of a Blackbox and constructing interpretable models. Beginning with a Blackbox, we iteratively carve out a mixture of interpretable experts (MoIE) and a residual network. Each interpretable model specializes in a subset of samples and explains them using First Order Logic (FOL), providing basic reasoning on concepts from the Blackbox. We route the remaining samples through a flexible residual. We repeat the method on the residual network until all the interpretable models explain the desired proportion of data. Our extensive experiments show that our route, interpret, and repeat approach (1) identifies a diverse set of instance-specific concepts with high concept completeness via MoIE without compromising in performance, (2) identifies the relatively ``harder'' samples to explain via residuals, (3) outperforms the interpretable by-design models by significant margins during test-time interventions, and (4) fixes the shortcut learned by the original Blackbox. The code for MoIE is publicly available at: \url{https://github.com/batmanlab/ICML-2023-Route-interpret-repeat}

Conventional physics-based modeling is a time-consuming bottleneck in control design for complex nonlinear systems like autonomous underwater vehicles (AUVs). In contrast, purely data-driven models, though convenient and quick to obtain, require a large number of observations and lack operational guarantees for safety-critical systems. Data-driven models leveraging available partially characterized dynamics have potential to provide reliable systems models in a typical data-limited scenario for high value complex systems, thereby avoiding months of expensive expert modeling time. In this work we explore this middle-ground between expert-modeled and pure data-driven modeling. We present control-oriented parametric models with varying levels of domain-awareness that exploit known system structure and prior physics knowledge to create constrained deep neural dynamical system models. We employ universal differential equations to construct data-driven blackbox and graybox representations of the AUV dynamics. In addition, we explore a hybrid formulation that explicitly models the residual error related to imperfect graybox models. We compare the prediction performance of the learned models for different distributions of initial conditions and control inputs to assess their accuracy, generalization, and suitability for control.

Artificial Intelligence (AI) is an increasing part of our everyday lives, powering our smartphones and the internet of things. But few people really understand what it is, how it works and more importantly, why it is so important to their business. The Oxford English Dictionary defines artificial intelligence (AI) as the theory and development of computer systems able to perform tasks normally requiring human intelligence such as visual perception, speech recognition, decision-making, and translation between languages. For many people in Business, the language used in data science can be confusing. It is far simpler to explain by simply saying, 'powered by AI'.

Informally, Adversarial ML is "subverting machine learning systems for fun and profit". The methods underpinning the production machine learning systems are systematically vulnerable to a new class of vulnerabilities across the machine learning supply chain collectively known as Adversarial Machine Learning. Adversaries can exploit these vulnerabilities to manipulate AI systems in order to alter their behavior to serve a malicious end goal. Consider a typical ML pipeline shown below that is gated behind an API, wherein the only way to use the model is to send a query and observe a response. In this example, we assume a blackbox setting: the attacker does NOT have direct access to the training data, no knowledge of the algorithm used and no source code of the model.

Artificial Intelligence innovation is stead fasting its progress across two segments, first, one being computer vision that has very nice applications of its own, the second one revolves around Natural Language Processing (NLP), which has quietly been gaining prominence credit to the popularity behind text analytics, speech recognition systems, becoming one of the most utilitarian tools for the enterprise today. Natural language technologies are a bouquet of three processes that explain AI's promise to let devices speak and understand the human language by combining the virtues of technology. Based on the ability of computers to run several algorithms to perform tasks like Automated Speech and Automated Text Writing in a short period, NLP is still considered the umbrella term that binds natural language technologies. Natural Language Understanding or NLU encapsulates to understand the meaning of the given text or word, its nature and structure and tries to resolve the ambiguity present in natural language which could be multiple meanings word (Lexical Ambiguity), sentences having multiple parse trees (Syntactic Ambiguity), and phrases or words which have been mentioned previously have a different meaning (Anaphoric Ambiguity). The final process of NLU involves understanding the meaning of each word by using lexicons (vocabulary) and a set of grammatical rules.

GBM models have been battle-tested as powerful models but have been tainted by the lack explainability. Typically data scientists look at variable importance plots but they are not enough to explain how a model works. To maximize adoption by the model user, use SHAP values to answer common explainability questions and build trust in your models. In this post, we will train a GBM model on a simple dataset and you will learn how to explain how the model works. The goal here is not to explain how the math works, but to explain to a non-technical user how the input variables are related to the output variable and how predictions are made.

We describe our submission to the Extreme Value Analysis 2019 Data Challenge in which teams were asked to predict extremes of sea surface temperature anomaly within spatio-temporal regions of missing data. We present a computational framework which reconstructs missing data using convolutional deep neural networks. Conditioned on incomplete data, we employ autoencoder-like models as multivariate conditional distributions from which possible reconstructions of the complete dataset are sampled using imputed noise. In order to mitigate bias introduced by any one particular model, a prediction ensemble is constructed to create the final distribution of extremal values. Our method does not rely on expert knowledge in order to accurately reproduce dynamic features of a complex oceanographic system with minimal assumptions. The obtained results promise reusability and generalization to other domains.