Neural Information Processing Systems http://nips.cc/

We thank the reviewers for acknowledging our contributions and providing valuable comments. We'll further improve the paper in the final version. We address the detail comments below. To R1: Q1: Relation with variants of DS: Our main goal is to provide a discriminative max-margin formulation, which is general and complementary to generative methods. For example, though we consider the vanilla DS in CrowdSVM for both clarity and space limit, other variants (e.g., [15,11]) can be naturally incorporated, as the RegBayes formulation (9) is generally applicable to any Bayesian models. Finally, the spectral initialization method [23] for confusion matrices can also be used to initialize the confusion matrices in CrowdSVM, so as the methods in [12].

Data-driven models are promising tools for predicting ocean conditions and enhancing the details of these predictions. In this study, we applied advanced machine learning methods to model sea surface velocity and height in the Gulf of Mexico. To forecast broad ocean conditions, we used a method called Fourier Neural Operators (FNO), designed to balance computational efficiency with accuracy through a specialized loss function that combines grid and spectral space information. For creating high-resolution details from low-resolution data -- a process called downscaling -- we explored two different neural network architectures and compared their performance against simpler linear interpolation. This combination of forecasting and downscaling methods greatly improves the efficiency of ocean forecast and downscaling compared to numerical simulation with limited input variables. Our results highlight that these data-driven techniques can provide reliable, physics-aware predictions that can be useful for quick, localized analyses and in generating statistical predictions.

Adaptive schemes, where tasks are assigned based on the data collected thus far, are widely used in practical crowdsourcing systems to efficiently allocate the budget. However, existing theoretical analyses of crowdsourcing systems suggest that the gain of adaptive task assignments is minimal. To bridge this gap, we investigate this question under a strictly more general probabilistic model, which has been recently introduced to model practical crowdsourcing datasets. Under this generalized Dawid-Skene model, we characterize the fundamental trade-off between budget and accuracy. We introduce a novel adaptive scheme that matches this fundamental limit. A given budget is allocated over multiple rounds. In each round, a subset of tasks with high enough confidence are classified, and increasing budget is allocated on remaining ones that are potentially more difficult. On each round, decisions are made based on the leading eigenvector of (weighted) non-backtracking operator corresponding to the bipartite assignment graph. We further quantify the gain of adaptivity, by comparing the tradeoff with the one for non-adaptive schemes, and confirm that the gain is significant and can be made arbitrarily large depending on the distribution of the difficulty level of the tasks at hand.

Dynamic Selection (DS), where base classifiers are chosen from a classifier's pool for each new instance at test time, has shown to be highly effective in pattern recognition. However, instability and redundancy in the classifier pools can impede computational efficiency and accuracy in dynamic ensemble selection. This paper introduces a meta-learning recommendation system (MLRS) to recommend the optimal pool generation scheme for DES methods tailored to individual datasets. The system employs a meta-model built from dataset meta-features to predict the most suitable pool generation scheme and DES method for a given dataset. Through an extensive experimental study encompassing 288 datasets, we demonstrate that this meta-learning recommendation system outperforms traditional fixed pool or DES method selection strategies, highlighting the efficacy of a meta-learning approach in refining DES method selection. The source code, datasets, and supplementary results can be found in this project's GitHub repository: https://github.com/Menelau/MLRS-PDS.

One of the primary catalysts fueling advances in artificial intelligence (AI) and machine learning (ML) is the availability of massive, curated datasets. A commonly used technique to curate such massive datasets is crowdsourcing, where data are dispatched to multiple annotators. The annotator-produced labels are then fused to serve downstream learning and inference tasks. This annotation process often creates noisy labels due to various reasons, such as the limited expertise, or unreliability of annotators, among others. Therefore, a core objective in crowdsourcing is to develop methods that effectively mitigate the negative impact of such label noise on learning tasks. This feature article introduces advances in learning from noisy crowdsourced labels. The focus is on key crowdsourcing models and their methodological treatments, from classical statistical models to recent deep learning-based approaches, emphasizing analytical insights and algorithmic developments. In particular, this article reviews the connections between signal processing (SP) theory and methods, such as identifiability of tensor and nonnegative matrix factorization, and novel, principled solutions of longstanding challenges in crowdsourcing -- showing how SP perspectives drive the advancements of this field. Furthermore, this article touches upon emerging topics that are critical for developing cutting-edge AI/ML systems, such as crowdsourcing in reinforcement learning with human feedback (RLHF) and direct preference optimization (DPO) that are key techniques for fine-tuning large language models (LLMs).

Drivers in ridesharing platforms exhibit cognitive atrophy and fatigue as they accept ride offers along the day, which can have a significant impact on the overall efficiency of the ridesharing platform. In contrast to the current literature which focuses primarily on modeling and learning driver's preferences across different ride offers, this paper proposes a novel Dynamic Discounted Satisficing (DDS) heuristic to model and predict driver's sequential ride decisions during a given shift. Based on DDS heuristic, a novel stochastic neural network with random activations is proposed to model DDS heuristic and predict the final decision made by a given driver. The presence of random activations in the network necessitated the development of a novel training algorithm called Sampling-Based Back Propagation Through Time (SBPTT), where gradients are computed for independent instances of neural networks (obtained via sampling the distribution of activation threshold) and aggregated to update the network parameters. Using both simulation experiments as well as on real Chicago taxi dataset, this paper demonstrates the improved performance of the proposed approach, when compared to state-of-the-art methods.

Remote sensing and artificial intelligence are pivotal technologies of precision agriculture nowadays. The efficient retrieval of large-scale field imagery combined with machine learning techniques shows success in various tasks like phenotyping, weeding, cropping, and disease control. This work will introduce a machine learning framework for automatized large-scale plant-specific trait annotation for the use case disease severity scoring for Cercospora Leaf Spot (CLS) in sugar beet. With concepts of Deep Label Distribution Learning (DLDL), special loss functions, and a tailored model architecture, we develop an efficient Vision Transformer based model for disease severity scoring called SugarViT. One novelty in this work is the combination of remote sensing data with environmental parameters of the experimental sites for disease severity prediction. Although the model is evaluated on this special use case, it is held as generic as possible to also be applicable to various image-based classification and regression tasks. With our framework, it is even possible to learn models on multi-objective problems as we show by a pretraining on environmental metadata.

While label fusion from multiple noisy annotations is a well understood concept in data wrangling (tackled for example by the Dawid-Skene (DS) model), we consider the extended problem of carrying out learning when the labels themselves are not consistently annotated with the same schema. We show that even if annotators use disparate, albeit related, label-sets, we can still draw inferences for the underlying full label-set. We propose the Inter-Schema AdapteR (ISAR) to translate the fully-specified label-set to the one used by each annotator, enabling learning under such heterogeneous schemas, without the need to re-annotate the data. We apply our method to a mouse behavioural dataset, achieving significant gains (compared with DS) in out-of-sample log-likelihood (-3.40 to -2.39) and F1-score (0.785 to 0.864).

Crowdsourcing is a strategy to categorize data through the contribution of many individuals. A wide range of theoretical and algorithmic contributions are based on the model of Dawid and Skene [1]. Recently it was shown in [2,3] that, in certain regimes, belief propagation is asymptotically optimal for data generated from the Dawid-Skene model. This paper is motivated by this recent progress. We analyze the dense limit of the Dawid-Skene model. It is shown that it belongs to a larger class of low-rank matrix estimation problems for which it is possible to express the asymptotic, Bayes-optimal, performance in a simple closed form. In the dense limit the mapping to a low-rank matrix estimation problem provides an approximate message passing algorithm that solves the problem algorithmically. We identify the regions where the algorithm efficiently computes the Bayes-optimal estimates. Our analysis refines the results of [2,3] about optimality of message passing algorithms by characterizing regions of parameters where these algorithms do not match the Bayes-optimal performance. We further study numerically the performance of approximate message passing, derived in the dense limit, on sparse instances and carry out experiments on a real world dataset.