arXiv research

This paper introduces $f$-DPO, a generalized approach to Direct Preference Optimization using diverse divergence constraints.

Master-slave architecture tackles combinatorial multi-armed bandits with diversity constraints.

In standard reinforcement learning (RL), a learning agent seeks to optimize the overall reward. However, many key aspects of a desired behavior are more naturally expressed as constraints. For instance, the designer may want to limit the use of unsafe actions, increase the diversity of trajectories to enable exploratio…

CPRA efficiently finds diverse solutions in CO problems using UL and parallelization.

New method improves text classification without labeled target data.

New method improves neural architecture search by optimizing for both performance and diversity.

Sampling methods that choose a subset of the data proportional to its diversity in the feature space are popular for data summarization. However, recent studies have noted the occurrence of bias (under- or over-representation of a certain gender or race) in such data summarization methods. In this paper we initiate a s…

Mixreg improves RL generalization by mixing diverse training environments.

PGEL learns embeddings to diversify protein motifs while maintaining biological function.

PhysVarMix predicts diverse urban trajectories with physics constraints.

We address the problem of partial index tracking, replicating a benchmark index using a small number of assets. Accurate tracking with a sparse portfolio is extensively studied as a classic finance problem. However in practice, a tracking portfolio must also be diverse in order to minimise risk -- a requirement which h…

Post-hoc explanations of machine learning models are crucial for people to understand and act on algorithmic predictions. An intriguing class of explanations is through counterfactuals, hypothetical examples that show people how to obtain a different prediction. We posit that effective counterfactual explanations shoul…

New algorithm clusters data and learns kernels without relaxing constraints.

δ-CLUE generates diverse explanations for model uncertainty.

Interactive steering improves hierarchical clustering for diverse user needs.

The broad set of deep generative models (DGMs) has achieved remarkable advances. However, it is often difficult to incorporate rich structured domain knowledge with the end-to-end DGMs. Posterior regularization (PR) offers a principled framework to impose structured constraints on probabilistic models, but has limited …

Paper tackles nonparametric classification with privacy constraints, achieving optimal accuracy.

In this paper we aim to find a measure for the diversity of cash flows between agents in an economy. We argue that cash flows can be linked to probabilities of finding a currency unit in a given cash flow. We then use the information entropy as a natural measure of diversity. This leads to a hirarchical inequality meas…

Study Federated RL with diverse constraints, proposing new optimization methods.

This paper proposes new search algorithms for counterfactual explanations based upon mixed integer programming. We are concerned with complex data in which variables may take any value from a contiguous range or an additional set of discrete states. We propose a novel set of constraints that we refer to as a "mixed pol…

Bayesian Entropy Neural Networks enforce constraints on deep learning predictions.

New algorithms for constrained online optimization with memory and predictions.

Optimising discrete data for a desired characteristic using gradient-based methods involves projecting the data into a continuous latent space and carrying out optimisation in this space. Carrying out global optimisation is difficult as optimisers are likely to follow gradients into regions of the latent space that the…

A new confidence measure improves self-training in biased data.

Paper proposes a risk-aware decision-making framework for real-world sequential decisions.

New method generates diverse EHR data types while maintaining privacy.

The paper studies constraint maps with singularities and free boundaries, proving continuity near singularities and optimality.

We propose a method to generate multiple diverse and valid human pose hypotheses in 3D all consistent with the 2D detection of joints in a monocular RGB image. We use a novel generative model uniform (unbiased) in the space of anatomically plausible 3D poses. Our model is compositional (produces a pose by combining par…

FedCONST adapts update magnitudes to enhance feature generalization in FL.

We describe society as a nonequilibrium probabilistic system: N individuals occupy W resource states in it and produce entropy S over definite time periods. Resulting thermodynamics is however unusual because a second entropy, H, measures a typically social feature, inequality or diversity in the distribution of availa…

New method robustly discovers causal relationships from imperfect data.

Generative model learns to create molecules with multiple properties using interpretable substructures.

The structure of complex networks has been of interest in many scientific and engineering disciplines over the decades. A number of studies in the field have been focused on finding the common properties among different kinds of networks such as heavy-tail degree distribution, small-worldness and modular structure and …

Framework optimizes model performance and interpretability for tabular data.

GACEM optimizes complex multi-modal problems using neural networks.

A new method combines SciML and UQ with physical constraints.

Binary hashing is a well-known approach for fast approximate nearest-neighbor search in information retrieval. Much work has focused on affinity-based objective functions involving the hash functions or binary codes. These objective functions encode neighborhood information between data points and are often inspired by…

New method generates equilibrium glass configurations efficiently.

Submodular functions are a broad class of set functions, which naturally arise in diverse areas. Many algorithms have been suggested for the maximization of these functions. Unfortunately, once the function deviates from submodularity, the known algorithms may perform arbitrarily poorly. Amending this issue, by obtaini…

New method identifies latent variables without strong assumptions.

Novel NAS method balances performance and hardware metrics efficiently.

This thesis improves protein contact prediction using unsupervised and supervised methods.

In many practical problems, a learning agent may want to learn the best action in hindsight without ever taking a bad action, which is significantly worse than the default production action. In general, this is impossible because the agent has to explore unknown actions, some of which can be bad, to learn better action…

We address the challenging problem of efficient inference across many devices and resource constraints, especially on edge devices. Conventional approaches either manually design or use neural architecture search (NAS) to find a specialized neural network and train it from scratch for each case, which is computationall…

L3Ms fine-tune LLMs with constraints for tailored applications.

MLDemon monitors ML systems post-deployment, improving reliability with real-time performance estimates and expert labels.

New method controls posterior collapse in VAEs without network architecture constraints.

Several fundamental problems that arise in optimization and computer science can be cast as follows: Given vectors $v_1,\ldots,v_m \in \mathbb{R}^d$ and a constraint family ${\cal B}\subseteq 2^{[m]}$, find a set $S \in \cal{B}$ that maximizes the squared volume of the simplex spanned by the vectors in $S$. A motivatin…