Pool-based Active Learning as Noisy Lossy Compression: Characterizing Label Complexity via Finite Blocklength Analysis

This paper proposes an information-theoretic framework for analyzing the theoretical limits of pool-based active learning (AL), in which a subset of instances is selectively labeled. The proposed framework reformulates pool-based AL as a noisy lossy compression problem by mapping pool observations to noisy symbol observations, data selection to compression, and learning to decoding. This correspondence enables a unified information-theoretic analysis of data selection and learning in pool-based AL. Applying finite blocklength analysis of noisy lossy compression, we derive information-theoretic lower bounds on label complexity and generalization error that serve as theoretical limits for a given learning algorithm under its associated optimal data selection strategy. Specifically, our bounds include terms that reflect overfitting induced by the learning algorithm and the discrepancy between its inductive bias and the target task, and are closely related to established information-theoretic bounds and stability theory, which have not been previously applied to the analysis of pool-based AL. These properties yield a new theoretical perspective on pool-based AL.

Supervised Learning as Lossy Compression: Characterizing Generalization and Sample Complexity via Finite Blocklength Analysis

This paper presents a novel information-theoretic perspective on generalization in machine learning by framing the learning problem within the context of lossy compression and applying finite blocklength analysis. In our approach, the sampling of training data formally corresponds to an encoding process, and the model construction to a decoding process. By leveraging finite blocklength analysis, we derive lower bounds on sample complexity and generalization error for a fixed randomized learning algorithm and its associated optimal sampling strategy. Our bounds explicitly characterize the degree of overfitting of the learning algorithm and the mismatch between its inductive bias and the task as distinct terms. This separation provides a significant advantage over existing frameworks. Additionally, we decompose the overfitting term to show its theoretical connection to existing metrics found in information-theoretic bounds and stability theory, unifying these perspectives under our proposed framework.

Learning from Similarity-Confidence and Confidence-Difference

In practical machine learning applications, it is often challenging to assign accurate labels to data, and increasing the number of labeled instances is often limited. In such cases, Weakly Supervised Learning (WSL), which enables training with incomplete or imprecise supervision, provides a practical and effective solution. However, most existing WSL methods focus on leveraging a single type of weak supervision. In this paper, we propose a novel WSL framework that leverages complementary weak supervision signals from multiple relational perspectives, which can be especially valuable when labeled data is limited. Specifically, we introduce SconfConfDiff Classification, a method that integrates two distinct forms of weaklabels: similarity-confidence and confidence-difference, which are assigned to unlabeled data pairs. To implement this method, we derive two types of unbiased risk estimators for classification: one based on a convex combination of existing estimators, and another newly designed by modeling the interaction between two weak labels. We prove that both estimators achieve optimal convergence rates with respect to estimation error bounds. Furthermore, we introduce a risk correction approach to mitigate overfitting caused by negative empirical risk, and provide theoretical analysis on the robustness of the proposed method against inaccurate class prior probability and label noise. Experimental results demonstrate that the proposed method consistently outperforms existing baselines across a variety of settings.

Learning from Hard Labels with Additional Supervision on Non-Hard-Labeled Classes

In scenarios where training data is limited due to observation costs or data scarcity, enriching the label information associated with each instance becomes crucial for building high-accuracy classification models. In such contexts, it is often feasible to obtain not only hard labels but also {\it additional supervision}, such as the confidences for the hard labels. This setting naturally raises fundamental questions: {\it What kinds of additional supervision are intrinsically beneficial?} And {\it how do they contribute to improved generalization performance?} To address these questions, we propose a theoretical framework that treats both hard labels and additional supervision as probability distributions, and constructs soft labels through their affine combination. Our theoretical analysis reveals that the essential component of additional supervision is not the confidence score of the assigned hard label, but rather the information of the distribution over the non-hard-labeled classes. Moreover, we demonstrate that the additional supervision and the mixing coefficient contribute to the refinement of soft labels in complementary roles. Intuitively, in the probability simplex, the additional supervision determines the direction in which the deterministic distribution representing the hard label should be adjusted toward the true label distribution, while the mixing coefficient controls the step size along that direction. Through generalization error analysis, we theoretically characterize how the additional supervision and its mixing coefficient affect both the convergence rate and asymptotic value of the error bound. Finally, we experimentally demonstrate that, based on our theory, designing additional supervision can lead to improved classification accuracy, even when utilized in a simple manner.

Learning from Complementary Features

While precise data observation is essential for the learning processes of predictive models, it can be challenging owing to factors such as insufficient observation accuracy, high collection costs, and privacy constraints. In this paper, we examines cases where some qualitative features are unavailable as precise information indicating "what it is," but rather as complementary information indicating "what it is not." We refer to features defined by precise information as ordinary features (OFs) and those defined by complementary information as complementary features (CFs). We then formulate a new learning scenario termed Complementary Feature Learning (CFL), where predictive models are constructed using instances consisting of OFs and CFs. The simplest formalization of CFL applies conventional supervised learning directly using the observed values of CFs. However, this approach does not resolve the ambiguity associated with CFs, making learning challenging and complicating the interpretation of the predictive model's specific predictions. Therefore, we derive an objective function from an information-theoretic perspective to estimate the OF values corresponding to CFs and to predict output labels based on these estimations. Based on this objective function, we propose a theoretically guaranteed graph-based estimation method along with its practical approximation, for estimating OF values corresponding to CFs. The results of numerical experiments conducted with real-world data demonstrate that our proposed method effectively estimates OF values corresponding to CFs and predicts output labels.