classification

This chapter reviews recent advances in weakly supervised learning, introducing confidence-difference classification, relaxed assumptions for complementary-label learning, and an evaluation framework for partial-label learning.

The paper introduces NysHD, a method that bridges hyperdimensional computing and kernel methods via the Nyström approximation, allowing any positive-semidefinite similarity function to be used as an HDC encoding. It demonstrates improved classification accuracy on graph and string datasets compared to existing HDC encoding methods.

This paper presents an automated item evaluation (AIE) model that predicts item acceptance/rejection in standardized testing using fine-tuned DeBERTa classifiers on raw item text and Qwen3-generated critiques, achieving moderate accuracy and highlighting challenges with fairness-related rejections.

This paper introduces Task-Conditional Flow Matching (TCFM), a framework for adapting multilingual text embedding models that selectively uses flow matching for translation tasks and other objectives for retrieval/classification. It achieves state-of-the-art results on the Indic Massive Text Embedding Benchmark.

This paper introduces MS-MLB, an open machine learning benchmark for classifying multiple sclerosis from whole blood RNA expression data using the public GSE17048 cohort. It provides a reproducible, leakage-controlled evaluation pipeline and reports Gradient Boosting as the top performer.

This paper introduces TS2TabPFN, a framework that combines explicit feature extraction with the TabPFN 2.5 tabular foundation model for time series classification and extrinsic regression. Experiments show it outperforms state-of-the-art models in TSER and achieves competitive results in TSC.

This paper introduces PatTree, a graph-based multimodal patient representation that automatically structures heterogeneous clinical data for medical classification tasks. It achieves state-of-the-art performance on the ADNI-1 cohort with 98.5% balanced accuracy for Alzheimer's disease classification.

The paper presents a methodology for building a large-scale Russian dataset from social media texts to detect presuicidal and anti-suicidal signals, including annotation guidelines and baseline classification experiments.

A developer shares lessons from building a production n8n AI support workflow, emphasizing classification over generation, rule-based filters to cut costs, and grounding responses in a knowledge base to avoid hallucinated answers.

This paper provides theoretical guarantees for distinguishing LLM responses by modeling token embeddings as trajectories of a dynamical system, proving exponential decay of misclassification probability and characterizing cross-embedding generalization.

This paper evaluates GRU, LSTM, and Transformer encoder models for classifying Level 2 automated driving systems from vehicle telematics data, and introduces a robustness evaluation framework simulating realistic data corruption.

A pre-registered ablation on a 4B model classification task shows that harness design (rule placement, evidence order, turn structure) swings accuracy from 60% to 82%, demonstrating that poor harness design is often mistaken for poor model capability.

An article classifying different approaches to using AI in software development, from traditional development to vibe coding, aiming to clarify debates by providing a shared vocabulary.

Pangram Labs presents Pangram 4, a state-of-the-art AI text detection model achieving 0.9916 AUROC with very low false positive and negative rates, and improved robustness to adversarial attacks and mixed authorship detection.

This paper presents a large-scale study comparing Bayesian complete-pooling models to frequentist baselines for cross-subject motor imagery EEG classification, finding that Bayesian methods improve reliability modestly but at a higher computational cost, with limited practical benefit.

AsymVerify is a confidence-gated verification system for political evasion detection, achieving 2nd place at SemEval-2026 Task 6 with 0.85 Macro F1. It uses an asymmetric two-verifier design to correct low-confidence predictions, improving over single-pass classification.

This paper explores using fine-tuned LLMs to identify indicators of vulnerability (mental ill health, substance misuse, alcohol dependence, homelessness) in UK police incident logs, finding that while LLMs can produce meaningful prevalence estimates, they require careful methodological support and are not reliable for individual-level decisions.

This paper introduces SEB-Cal, a method that augments output-space calibration with spectral features (band energy, entropy, peak dominance, phase stability) to improve selective reliability estimation in time-series classification, achieving higher Corr-AUROC and lower [email protected] across multiple datasets.

This paper presents a lightweight 1D CNN framework for classifying affective touches in soft plush companions, achieving 75% test accuracy with a 13.2k-parameter model and proposing a hybrid heuristic-CNN inference pipeline for embedded deployment.

This paper presents a comprehensive comparison of deep learning architectures, including Vision Transformers and Graph Attention Networks, for automated sleep apnea detection from multichannel EEG signals, achieving a best test AUC of 0.750 using a vision transformer trained on topological data analysis features.