Variance-Aware Loss Scheduling for Multimodal Alignment in Low-Data Settings

Training vision-language models for image-text alignment typically requires large datasets to achieve robust performance. In low-data scenarios, standard contrastive learning can struggle to align modalities effectively due to overfitting and unstable training dynamics. In this paper, we propose a variance-aware loss scheduling approach that dynamically adjusts the weighting of the contrastive loss based on the statistical variability (uncertainty) in the model's alignment predictions. Using a subset of the Flickr8k image-caption dataset to simulate limited data conditions, we demonstrate that our approach improves image-text retrieval accuracy compared to a fixed-weight baseline. We also compare against other adaptive weighting strategies (using output entropy and cosine similarity spread) and find that variance-aware scheduling provides the best overall trade-off. Qualitatively, our method yields more distinct multimodal embeddings as shown by t-SNE visualizations. Moreover, in a stress test with noise-injected captions and images, the variance-guided loss proves more robust, maintaining higher recall when random perturbations are introduced. These results highlight the benefit of adaptive loss weighting for multimodal alignment in low-data regimes.

Learning Data-Driven Uncertainty Set Partitions for Robust and Adaptive Energy Forecasting with Missing Data

Short-term forecasting models typically assume the availability of input data (features) when they are deployed and in use. However, equipment failures, disruptions, cyberattacks, may lead to missing features when such models are used operationally, which could negatively affect forecast accuracy, and result in suboptimal operational decisions. In this paper, we use adaptive robust optimization and adversarial machine learning to develop forecasting models that seamlessly handle missing data operationally. We propose linear- and neural network-based forecasting models with parameters that adapt to available features, combining linear adaptation with a novel algorithm for learning data-driven uncertainty set partitions. The proposed adaptive models do not rely on identifying historical missing data patterns and are suitable for real-time operations under stringent time constraints. Extensive numerical experiments on short-term wind power forecasting considering horizons from 15 minutes to 4 hours ahead illustrate that our proposed adaptive models are on par with imputation when data are missing for very short periods (e.g., when only the latest measurement is missing) whereas they significantly outperform imputation when data are missing for longer periods. We further provide insights by showcasing how linear adaptation and data-driven partitions (even with a few subsets) approach the performance of the optimal, yet impractical, method of retraining for every possible realization of missing data.

From Data to Uncertainty Sets: a Machine Learning Approach

Existing approaches of prescriptive analytics -- where inputs of an optimization model can be predicted by leveraging covariates in a machine learning model -- often attempt to optimize the mean value of an uncertain objective. However, when applied to uncertain constraints, these methods rarely work because satisfying a crucial constraint in expectation may result in a high probability of violation. To remedy this, we leverage robust optimization to protect a constraint against the uncertainty of a machine learning model's output. To do so, we design an uncertainty set based on the model's loss function. Intuitively, this approach attempts to minimize the uncertainty around a prediction. Extending guarantees from the robust optimization literature, we derive strong guarantees on the probability of violation. On synthetic computational experiments, our method requires uncertainty sets with radii up to one order of magnitude smaller than those of other approaches.

Core-Set Selection for Data-efficient Land Cover Segmentation

The increasing accessibility of remotely sensed data and the potential of such data to inform large-scale decision-making has driven the development of deep learning models for many Earth Observation tasks. Traditionally, such models must be trained on large datasets. However, the common assumption that broadly larger datasets lead to better outcomes tends to overlook the complexities of the data distribution, the potential for introducing biases and noise, and the computational resources required for processing and storing vast datasets. Therefore, effective solutions should consider both the quantity and quality of data. In this paper, we propose six novel core-set selection methods for selecting important subsets of samples from remote sensing image segmentation datasets that rely on imagery only, labels only, and a combination of each. We benchmark these approaches against a random-selection baseline on three commonly used land cover classification datasets: DFC2022, Vaihingen, and Potsdam. In each of the datasets, we demonstrate that training on a subset of samples outperforms the random baseline, and some approaches outperform training on all available data. This result shows the importance and potential of data-centric learning for the remote sensing domain. The code is available at https://github.com/keillernogueira/data-centric-rs-classification/.

Data-Driven Calibration of Prediction Sets in Large Vision-Language Models Based on Inductive Conformal Prediction

This study addresses the critical challenge of hallucination mitigation in Large Vision-Language Models (LVLMs) for Visual Question Answering (VQA) tasks through a Split Conformal Prediction (SCP) framework. While LVLMs excel in multi-modal reasoning, their outputs often exhibit hallucinated content with high confidence, posing risks in safety-critical applications. We propose a model-agnostic uncertainty quantification method that integrates dynamic threshold calibration and cross-modal consistency verification. By partitioning data into calibration and test sets, the framework computes nonconformity scores to construct prediction sets with statistical guarantees under user-defined risk levels ($\alpha$). Key innovations include: (1) rigorous control of \textbf{marginal coverage} to ensure empirical error rates remain strictly below $\alpha$; (2) dynamic adjustment of prediction set sizes inversely with $\alpha$, filtering low-confidence outputs; (3) elimination of prior distribution assumptions and retraining requirements. Evaluations on benchmarks (ScienceQA, MMMU) with eight LVLMs demonstrate that SCP enforces theoretical guarantees across all $\alpha$ values. The framework achieves stable performance across varying calibration-to-test split ratios, underscoring its robustness for real-world deployment in healthcare, autonomous systems, and other safety-sensitive domains. This work bridges the gap between theoretical reliability and practical applicability in multi-modal AI systems, offering a scalable solution for hallucination detection and uncertainty-aware decision-making.

A Novel Decomposed Feature-Oriented Framework for Open-Set Semantic Segmentation on LiDAR Data

Semantic segmentation is a key technique that enables mobile robots to understand and navigate surrounding environments autonomously. However, most existing works focus on segmenting known objects, overlooking the identification of unknown classes, which is common in real-world applications. In this paper, we propose a feature-oriented framework for open-set semantic segmentation on LiDAR data, capable of identifying unknown objects while retaining the ability to classify known ones. We design a decomposed dual-decoder network to simultaneously perform closed-set semantic segmentation and generate distinctive features for unknown objects. The network is trained with multi-objective loss functions to capture the characteristics of known and unknown objects. Using the extracted features, we introduce an anomaly detection mechanism to identify unknown objects. By integrating the results of close-set semantic segmentation and anomaly detection, we achieve effective feature-driven LiDAR open-set semantic segmentation. Evaluations on both SemanticKITTI and nuScenes datasets demonstrate that our proposed framework significantly outperforms state-of-the-art methods. The source code will be made publicly available at https://github.com/nubot-nudt/DOSS.

Image Data Augmentation for the TAIGA-IACT Experiment with Conditional Generative Adversarial Networks

Modern Imaging Atmospheric Cherenkov Telescopes (IACTs) generate a huge amount of data that must be classified automatically, ideally in real time. Currently, machine learning-based solutions are increasingly being used to solve classification problems. However, these classifiers require proper training data sets to work correctly. The problem with training neural networks on real IACT data is that these data need to be pre-labeled, whereas such labeling is difficult and its results are estimates. In addition, the distribution of incoming events is highly imbalanced. Firstly, there is an imbalance in the types of events, since the number of detected gamma quanta is significantly less than the number of protons. Secondly, the energy distribution of particles of the same type is also imbalanced, since high-energy particles are extremely rare. This imbalance results in poorly trained classifiers that, once trained, do not handle rare events correctly. Using only conventional Monte Carlo event simulation methods to solve this problem is possible, but extremely resource-intensive and time-consuming. To address this issue, we propose to perform data augmentation with artificially generated events of the desired type and energy using conditional generative adversarial networks (cGANs), distinguishing classes by energy values. In the paper, we describe a simple algorithm for generating balanced data sets using cGANs. Thus, the proposed neural network model produces both imbalanced data sets for physical analysis as well as balanced data sets suitable for training other neural networks.

Deep Diffusion Maps

One of the fundamental problems within the field of machine learning is dimensionality reduction. Dimensionality reduction methods make it possible to combat the so-called curse of dimensionality, visualize high-dimensional data and, in general, improve the efficiency of storing and processing large data sets. One of the best-known nonlinear dimensionality reduction methods is Diffusion Maps. However, despite their virtues, both Diffusion Maps and many other manifold learning methods based on the spectral decomposition of kernel matrices have drawbacks such as the inability to apply them to data outside the initial set, their computational complexity, and high memory costs for large data sets. In this work, we propose to alleviate these problems by resorting to deep learning. Specifically, a new formulation of Diffusion Maps embedding is offered as a solution to a certain unconstrained minimization problem and, based on it, a cost function to train a neural network which computes Diffusion Maps embedding -- both inside and outside the training sample -- without the need to perform any spectral decomposition. The capabilities of this approach are compared on different data sets, both real and synthetic, with those of Diffusion Maps and the Nystrom method.

Scalable Substructure Discovery Algorithm For Homogeneous Multilayer Networks

Graph mining analyzes real-world graphs to find core substructures (connected subgraphs) in applications modeled as graphs. Substructure discovery is a process that involves identifying meaningful patterns, structures, or components within a large data set. These substructures can be of various types, such as frequent patterns, motifs, or other relevant features within the data. To model complex data sets -- with multiple types of entities and relationships -- multilayer networks (or MLNs) have been shown to be more effective as compared to simple and attributed graphs. Analysis algorithms on MLNs using the decoupling approach have been shown to be both efficient and accurate. Hence, this paper focuses on substructure discovery in homogeneous multilayer networks (one type of MLN) using a novel decoupling-based approach. In this approach, each layer is processed independently, and then the results from two or more layers are composed to identify substructures in the entire MLN. The algorithm is designed and implemented, including the composition part, using one of the distributed processing frameworks (the Map/Reduce paradigm) to provide scalability. After establishing the correctness, we analyze the speedup and response time of the proposed algorithm and approach through extensive experimental analysis on large synthetic and real-world data sets with diverse graph characteristics.

ACE: A Cardinality Estimator for Set-Valued Queries

Cardinality estimation is a fundamental functionality in database systems. Most existing cardinality estimators focus on handling predicates over numeric or categorical data. They have largely omitted an important data type, set-valued data, which frequently occur in contemporary applications such as information retrieval and recommender systems. The few existing estimators for such data either favor high-frequency elements or rely on a partial independence assumption, which limits their practical applicability. We propose ACE, an Attention-based Cardinality Estimator for estimating the cardinality of queries over set-valued data. We first design a distillation-based data encoder to condense the dataset into a compact matrix. We then design an attention-based query analyzer to capture correlations among query elements. To handle variable-sized queries, a pooling module is introduced, followed by a regression model (MLP) to generate final cardinality estimates. We evaluate ACE on three datasets with varying query element distributions, demonstrating that ACE outperforms the state-of-the-art competitors in terms of both accuracy and efficiency.