BackboneLearn speeds up MIO-based machine learning problems.
problem Scaling mixed-integer optimization problems in machine learning.
method An open-source Python framework for MIO problems with indicator variables.
result Solves MIO problems faster and more accurately than existing methods.
In the last twenty-five years (1990-2014), algorithmic advances in integer optimization combined with hardware improvements have resulted in an astonishing 200 billion factor speedup in solving Mixed Integer Optimization (MIO) problems. We present a MIO approach for solving the classical best subset selection problem o…
Unified framework for intersectionally fair AI models using MIO.
problem Bias in AI models for high-risk domains.
method Mixed-Integer Optimization (MIO) for fairness and interpretability.
result Improved performance in detecting and mitigating bias at intersections.
EBBS integrates expert assessments into MIO best-subsets problem.
problem Identifying relevant features in statistical models.
method Expert-implied Bayesian approach using MIO.
result Reduces to classical best-subsets when experts are silent.
The Bryant-Ferry-Mio-Weinberger surgery exact sequence for high-dimensional compact ANR homology manifolds is used to obtain transversality, splitting and bordism results for homology manifolds, generalizing previous work of Johnston.
This paper uses MIO to select features for kernel SVM classification.
problem Feature selection for kernel SVM classification.
method Mixed-integer optimization (MIO) for feature subset selection.
result The MIO approach can often outperform linear-SVM-based methods in prediction performance.
New method optimizes mixed integer optimization for hierarchical modeling of clustered and longitudinal data.
problem Optimizing subset selection in hierarchical models with clustered and longitudinal data.
method Distribution-free mixed-integer optimization approach for cluster-aware regression.
result The method efficiently solves problems within minutes and outperforms traditional models in generating sparse solutions with high predictive power.
An explicit (-1)^n-quadratic form over Z[Z^{2n}] representing the surgery problem E_8 x T^{2n} is obtained, for use in the Bryant-Ferry-Mio-Weinberger construction of 2n-dimensional exotic homology manifolds.
Mixed-integer optimization improves fairness and transparency in machine learning models.
problem Ensuring fairness and transparency in machine learning models deployed in sensitive areas.
method Embedding responsible ML considerations directly into the learning process using mixed-integer optimization.
result MIO enables the learning of inherently transparent models that can incorporate fairness or other constraints.
We provide a proof of the controlled surgery sequence, including stability, in the special case that the local fundamental groups are trivial. Stability is a key ingredient in the construction of exotic homology manifolds by Bryant, Ferry, Mio and Weinberger, but no proof has been available. The development given here …
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…
New insights link RLHF and contrastive learning for better model alignment.
problem Aligning large language models with human values.
method Interpreting RLHF and DPO as contrastive learning methods based on mutual information.
result Proposed Mutual Information Optimization (MIO) improves model performance.
A new combinatorial approach groups regression coefficients for improved accuracy.
problem Grouping regression coefficients to reveal shared values within groups.
method Introduces L0-Fusion, a combinatorial grouping approach using mixed integer optimization. result L0-Fusion achieves grouping consistency under weak grouping sensitivity conditions. Optimal Survival Trees improve accuracy in medical data analysis.
problem Analyzing censored outcomes in medical data.
method Mixed-integer optimization and local search techniques.
result Improves accuracy in large datasets compared to existing methods.
The L0-regularized least squares problem (a.k.a. best subsets) is central to sparse statistical learning and has attracted significant attention across the wider statistics, machine learning, and optimization communities. Recent work has shown that modern mixed integer optimization (MIO) solvers can be used to addre…
Contradicts claims about Poincaré complexes and homology manifolds.
problem Claims about Poincaré complexes and homology manifolds are contradicted.
method Constructs a Poincaré complex with specific properties to contradict the claims.
result A Poincaré complex with vanishing periodic total surgery obstruction is not necessarily homotopy equivalent to a homology manifold.
AF improves classification models by adaptively weighting trees.
problem Improving classification model performance.
method AF combines OP2T for input-dependent weights and MIO for dynamic refinement.
result AF consistently outperforms RF, XGBoost, and other weighted RF.
We study the pull-back of the 2-parameter family of quotient elastic metrics introduced in Mio-Srivastava-Joshi on the space of arc-length parameterized loops. This point of view has the advantage of concentrating on the manifold of arc-length parameterized curves, which is a very natural manifold when the analysis of …
ODTLearn learns optimal decision trees for predictive and prescriptive tasks.
problem Learning optimal decision trees for high-stakes predictive and prescriptive tasks.
method Mixed-integer optimization framework and object-oriented design.
result Implementation of optimal decision trees for various tasks.
Aave community attack led to irretrievable debt and questioned decentralization in DeFi lending.
problem Decentralization vs. security in large DeFi lending protocols.
method Analyzed Avi Eisenberg's attack on Aave, showing how he artificially deflated CRV price.
result Attack led to irretrievable debt of over $1.5 million and highlighted decentralization vs. security.
Method learns optimal treatment policies from observational data.
problem Learning interpretable treatment assignment policies from observational data.
method Mixed-integer optimization (MIO) technology.
result Asymptotically exact in converging to optimal treatment policies.
Occupant behavior (OB) and in particular window openings need to be considered in building performance simulation (BPS), in order to realistically model the indoor climate and energy consumption for heating ventilation and air conditioning (HVAC). However, the proposed OB window opening models are often biased towards …
New method estimates sparse covariance matrices in logit mixtures.
problem Estimating correlations among random coefficients in logit models.
method Mixed-integer optimization (MIO) with Markov Chain Monte Carlo (MCMC) for posterior draws.
result Correctly recovers true covariance structure from synthetic data.
Proposes a new method for subgroup analysis using optimal trees with parameter fusion.
problem Challenges of greedy heuristics and overfitting in tree-based recursive partitioning methods.
method Fused optimal causal tree method leveraging mixed integer optimization (MIO) for globally optimal partitions and parameter fusion.
result Substantial improvement in subgroup discovery accuracy and statistical efficiency.
The primary purpose of this paper concerns the relation of (compact) generalized manifolds to finite Poincaré duality complexes (PD complexes). The problem is that an arbitrary generalized manifold X is always an ENR space, but it is not necessarily a complex. Moreover, finite PD complexes require the Poincaré dualit…
A new metric MSD detects bias in datasets efficiently.
problem Detecting bias in AI systems and datasets.
method Introduced Maximum Subgroup Discrepancy (MSD) metric and a practical algorithm based on MIO.
result MSD provides a linear sample complexity for practical applications, distinguishing biases effectively.
A new mutual information optimization method using self-supervised binary contrastive learning.
problem Improving self-supervised contrastive learning for better model performance.
method Proposes a novel loss function for contrastive learning that optimizes mutual information in positive and negative pairs.
result The proposed method outperforms state-of-the-art self-supervised contrastive frameworks on various benchmark datasets.
The study of Reeb dynamics on contact manifolds without periodic orbits.
problem Existence of periodic Reeb orbits on bm-contact manifolds. method Generalization of the Weinstein conjecture, proof of periodic orbits, existence of traps.
result In dimension 3, there are infinitely many periodic orbits on the critical set.
DeFi lending protocols faced challenges during Ethereum's merge, but avoided major liquidations.
problem Ethereum's merge caused volatility and potential liquidations in DeFi lending.
method Analyzed AAVE and Compound lending protocols during the merge and hard fork.
result Borrowing rates spiked but no significant liquidations occurred.
We show that an n-dimensional compactum X embeds in R^m, where m>3(n+1)/2, if and only if X x X - Δadmits an equivariant map to S^{m-1}. In particular, X embeds in R^{2n}, n>3, iff the top power of the (twisted) Euler class of the factor-exchanging involution on X x X - Δis trivial. Assuming that X quasi-embeds in R^{2…