Physics analogies explain machine learning overfitting control.
problem Understanding and controlling overfitting in machine learning.
method Analogies from physics and biology applied to algorithmic stability and GAN models.
result Physics formulas provide insights into reducing overfitting in machine learning.
Deep neural network learns optimal trading controls for high-frequency finance.
problem Optimal trading on high-frequency data with market impact and limited data.
method Deep neural network, Monte-Carlo initialization, transfer learning, explainable controls.
result Neural network learns optimal controls for trader preferences.
The aim of this paper is to explain how parameters adjustments can be integrated in the design or the control of automates of trading. Typically, we are interested by the online estimation of the market impacts generated by robots or single orders, and how they/the controller should react in an optimal way to the infor…
We explain the geometric origin of the L∞-algebra controlling deformations of pre-symplectic structures.
Unified Bayesian model explains in-context learning and activation steering in LLMs.
problem Understanding and controlling the behavior of large language models (LLMs) through prompts and activations.
method Developed a Bayesian model to explain and predict the effects of in-context learning and activation steering.
result Unified model predicts distinct phases and sudden shifts in LLM behavior, explaining prior empirical phenomena.
Geometric framework explains and controls implicit bias in machine learning.
problem Understanding and controlling the selection of solutions in overparameterized models.
method Developed a theoretical and constructive framework based on geometric corrections induced by gradient noise and continuous symmetries of the loss.
result Computed the induced bias across various architectures and enabled inverse design to shape the bias.
Proposes a method for explaining tabular data using copulas.
problem Lack of ground truth for explainability in complex datasets.
method Uses copulas to specify statistical properties and build intuition.
result Demonstrates improved explainability on logistic regression and correlation use cases.
CLASSIX is a fast and explainable clustering method that sorts data and merges groups.
problem Clustering of data with various shapes and dimensions.
method Greedy aggregation followed by cluster merging with scalar parameters.
result CLASSIX performs competitively with state-of-the-art algorithms and provides intuitive explanations.
Paper interprets ResNets via gate-network controls and deep-layer classifications.
problem Understanding the performance mechanism of ResNets.
method Constructs typical solutions using gate-network controls and deep-layer classifications.
result Proves the universal-approximation capability of ResNets.
We adapt Shapley values to explain model uncertainty, connecting it to information theory.
problem Explaining uncertainty in model predictions.
method Adapted Shapley value framework to quantify feature contributions to predictive uncertainty.
result Deep connections between Shapley values and information theory quantities.
Data science principles enhance AI interpretability for better user control.
problem Risks from opaque AI models without clear impacts.
method Synthesizes principles from interpretability literature, emphasizing audience goals.
result Illustrates basic techniques and criteria for evaluating interpretability.
Abstract: Surveying connections between ML and Control Theory.
problem Addressing the intersection of Machine Learning and Control Theory.
method Develops connections through reinforcement learning, supervised learning, deep learning, and stochastic gradient descent.
result Machine Learning and Control Theory are interconnected, with ML solving large control problems and Control Theory providing tools for ML.
ControlSHAP stabilizes Shapley value approximations using control variates.
problem High computational cost of exact Shapley values in blackbox models.
method ControlSHAP uses Monte Carlo control variates to stabilize Shapley value approximations.
result Significant reduction in Monte Carlo variability of Shapley estimates.
New framework for explainable AI on high-dimensional data.
problem Challenges in explainability with high-dimensional data.
method Two modules: latent representation and Shapley paradigm adaptation.
result Interpretable model explanations for high-dimensional data.
Sparse PCA selects variables with FDR control for improved performance.
problem Sparse PCA selects irrelevant variables when maximizing explained variance.
method Proposes FDR-controlled selection using T-Rex selector.
result Significant performance improvement over traditional sparse PCA.
Explainable AI improves human decision accuracy but does not enhance it significantly.
problem Improving human decision-making through explainable AI.
method Comparing human decision accuracy with and without AI predictions, including or excluding explanations.
result Providing AI predictions improves human decision accuracy, but explanations do not significantly enhance it.
Study uses multi-agent reinforcement learning to control self-assembly with high-resolution external control.
problem Designing effective external control protocols for self-assembly with high-resolution control.
method Investigated a multi-agent reinforcement learning approach, comparing fully decentralized and partially decentralized strategies.
result Partially decentralized approach outperforms fully decentralized in controlling self-assembly towards target structures.
Extremely accurate prediction of dynamical system bifurcations using control inputs.
problem Predicting complex bifurcation structures in dynamical systems.
method Extending extreme learning machines with control inputs to model system dynamics.
result The model can nearly reproduce the entire structure of bifurcations using only a few parameter values.
This paper uses NLDT to find interpretable control rules from complex DRL policies.
problem Complex, non-interpretable policies from black-box AI methods.
method Evolutionary optimization of NLDT for hierarchical control rules.
result Interpretable control rules with similar performance to black-box DRL.
CoxSE combines deep learning with self-explaining neural networks for survival analysis.
problem Improving predictive power of Cox Proportional Hazards model while maintaining explainability.
method Proposes CoxSE, a locally explainable Cox proportional hazards model using SENN, and CoxSENAM, a hybrid model with NAM.
result CoxSE provides more stable and consistent explanations while maintaining predictive power.
LLMs can help explain credit risk models but not autonomously.
problem Leveraging LLMs for post-hoc explainability in credit risk models.
method Comparison of LLM outputs with SHAP and coefficient-based attributions on three LMs.
result LLMs reliably preserve feature-importance rankings but poorly align with autonomous explanations.
New method for explaining neural network activation functions.
problem Transparency in black-box deep learning algorithms.
method Symbolic explanation of activation functions using adaptive Gaussian Processes.
result Achieved partially explainable learning model with scalable topology.
msPCA solves sparse PCA for multiple components efficiently.
problem Sparse principal component analysis with multiple components.
method Alternating maximization algorithm for sparse loading vectors, with orthogonality or zero correlation constraints.
result Achieves high variance explained with sparse components and controlled feasibility violations.
ALMANACS benchmarks explainability methods on simulatability.
problem Evaluating the effectiveness of explainability methods for language models.
method ALMANACS is a simulatability benchmark that evaluates explainability methods on twelve safety-relevant topics.
result No explainability method outperforms the explanation-free control across all topics.
Study proposes a data-driven CBR system for improved bankruptcy prediction.
problem Lack of interpretability in machine learning models for bankruptcy prediction.
method Data-driven explainable case-based reasoning (CBR) system.
result Proposed CBR system outperforms existing CBR and machine learning models.
Decision tree learning is a popular classification technique most commonly used in machine learning applications. Recent work has shown that decision trees can be used to represent provably-correct controllers concisely. Compared to representations using lookup tables or binary decision diagrams, decision trees are sma…
Survey on combining causal models with deep generative models for improved explainability and fairness.
problem Deep generative models lack explainability, induce spurious correlations, and poor out-of-distribution extrapolation.
method Structural causal models (SCMs) combined with deep generative models to address shortcomings.
result Causal generative models offer robustness, fairness, and interpretability.
Explearn learns to explain predictions using Gaussian Processes.
problem Learning to explain predictions effectively.
method Gaussian Processes-based contextual bandits.
result Guaranteed convergence with high probability.
We define a new condition on relatively hyperbolic Dehn filling which allows us to control the behavior of a relatively quasiconvex subgroups which need not be full. As an application, in combination with a recent result of Cooper and Futer, we provide a new proof of the virtual fibering of non-compact finite-volume hy…
Sparse text alignments learned via optimal transport improve model explainability.
problem Building self-explaining models by selecting relevant text pieces.
method Employing optimal transport to find minimal cost alignments, introducing constrained variants for sparsity.
result Sparse and interpretable alignments achieved, preserving prediction accuracy.
A novel controller for wheeled robots handles joystick inputs for smooth steering.
problem Steering control for differential-drive wheeled robots from indirect joystick inputs.
method Developed a geometric controller based on Darboux frame kinematics.
result Smooth trajectories achieved with safety constraints and no desired states.
CUBE explains models by balanced experiments and contrasts.
problem Post-hoc explanation of trained predictive models.
method Design-based framework using balanced low-high probes.
result Reveals dominant learned effect structure and clarifies query efficiency.
The Interactive Minority Game (IMG) is an online version of the traditional Minority Game in which human players can enter into competition with the traditional computer-controlled agents. Through the rich (and, importantly, analytically understood) behaviour of the MG, we can explore humans' behaviour in different kin…
Deep residual networks can approximate any continuous function using control theory.
problem Universal approximation capabilities of deep residual neural networks.
method Relating residual networks to control systems and using Lie algebraic techniques.
result Deep residual networks with adequately deep layers can approximate any continuous function on a compact set.
Following Bryant, Ferry, Mio and Weinberger we construct generalized manifolds as limits of controlled sequences p_i: X_i --> X_{i-1} : i = 1,2,... of controlled Poincaré spaces. The basic ingredient is the epsilon-delta-surgery sequence recently proved by Pedersen, Quinn and Ranicki. Since one has to apply it not only…
This paper develops explainable treatment policies for RPM using clinical knowledge.
problem Barriers to adoption of DHIs and lack of interpretability in purely black-box algorithms.
method Developed a pipeline for learning explainable treatment policies using clinician-informed representations.
result Policies learned from clinician-informed representations are more efficacious and efficient than black-box policies.
Clarifies relation for solving control-affine Schrödinger bridge problems.
problem Solving control-affine Schrödinger bridge problems via Hopf-Cole transform.
method Applies Hopf-Cole transform to conditions of optimality, resulting in nonlinear PDEs.
result Generic control-affine Schrödinger bridge requires further algorithmic development.
Recently, deep learning has been advancing the state of the art in artificial intelligence to a new level, and humans rely on artificial intelligence techniques more than ever. However, even with such unprecedented advancements, the lack of explanation regarding the decisions made by deep learning models and absence of…
This survey reviews portfolio choice in settings where investment opportunities are stochastic due to, e.g., stochastic volatility or return predictability. It is explained how to heuristically compute candidate optimal portfolios using tools from stochastic control, and how to rigorously verify their optimality by mea…
Nonlinear optimal control problems are often solved with numerical methods that require knowledge of system's dynamics which may be difficult to infer, and that carry a large computational cost associated with iterative calculations. We present a novel neurobiologically inspired hierarchical learning framework, Reinfor…
In this paper we present the solution to a longstanding problem of differential geometry: Lie's third theorem for Lie algebroids. We show that the integrability problem is controlled by two computable obstructions. As applications we derive, explain and improve the known integrability results, we establish integrabilit…
Schrödinger bridge solved with Weyl calculus for quadratic state cost.
problem Optimal control policy to steer joint state statistics.
method Weyl calculus in quantum mechanics for reaction-diffusion PDEs.
result Explicit Markov kernel for quadratic state cost found.
This work explains RL policies using causal models, revealing important patterns and failures.
problem Understanding why RL policies succeed or fail in complex, high-dimensional systems.
method Developed a nonlinear Causal Model Reduction framework to learn simplified causal models from RL policy actions and rewards.
result The approach can uncover important behavioral patterns and failure modes in trained RL policies.
Evolutionary methods improve understanding of LLMs and their relationships.
problem Improving understanding of LLMs and their relationships.
method Relating weights to genotypes and output text to phenotypes using evolutionary methods.
result Estimated evolutionary trees reliably recover the topology of the ground-truth training tree.
Spectral normalization stabilizes GANs by controlling gradient explosion and vanishing.
problem Stability and sample quality issues in GAN training.
method Spectral normalization controls gradient explosion and vanishing, improving GAN training stability and sample quality.
result Bidirectional Scaled Spectral Normalization (BSSN) outperforms standard spectral normalization in sample quality and training stability.
As machine learning becomes an important part of many real world applications affecting human lives, new requirements, besides high predictive accuracy, become important. One important requirement is transparency, which has been associated with model interpretability. Many machine learning algorithms induce models diff…
This paper describes Simpson's paradox, and explains its serious implications for randomised control trials. In particular, we show that for any number of variables we can simulate the result of a controlled trial which uniformly points to one conclusion (such as 'drug is effective') for every possible combination of t…
In this paper we propose a mathematical framework to address the uncertainty emergingwhen the designer of a trading algorithm uses a threshold on a signal as a control. We rely ona theorem by Benveniste and Priouret to deduce our Inventory Asymptotic Behaviour (IAB)Theorem giving the full distribution of the inventory …