Search for: All records

Mitigating label noise is a crucial problem in classification. Noise filtering is an effective method of dealing with label noise which does not need to estimate the noise rate or rely on any loss function. However, most filtering methods focus mainly on binary classification, leaving the more difficult counterpart problem of multiclass classification relatively unexplored. To remedy this deficit, we present a definition for label noise in a multiclass setting and propose a general framework for a novel label noise filtering learning method for multiclass classification. Two examples of noise filtering methods for multiclass classification, multiclass complete random forest (mCRF) and multiclass relative density, are derived from their binary counterparts using our proposed framework. In addition, to optimize the NI_threshold hyperparameter in mCRF, we propose two new optimization methods: a new voting cross-validation method and an adaptive method that employs a 2-means clustering algorithm. Furthermore, we incorporate SMOTE into our label noise filtering learning framework to handle the ubiquitous problem of imbalanced data in multiclass classification. We report experiments on both synthetic data sets and UCI benchmarks to demonstrate our proposed methods are highly robust to label noise in comparison with state-of-the-art baselines. All code and data results are available at https://github.com/syxiaa/Multiclass-Label-Noise-Filtering-Learning. 

Alzheimer's disease (AD) is a neurodegenerative disorder with slow onset, which could result in the deterioration of the duration of persistent neurological dysfunction. How to identify the informative longitudinal phenotypic neuroimaging markers and predict cognitive measures are crucial to recognize AD at early stage. Many existing models related imaging measures to cognitive status using regression models, but they did not take full consideration of the interaction between cognitive scores. In this paper, we propose a robust low-rank structured sparse regression method (RLSR) to address this issue. The proposed model simultaneously selects effective features and learns the underlying structure between cognitive scores by utilizing novel mixed structured sparsity inducing norms and low-rank approximation. In addition, an efficient algorithm is derived to solve the proposed non-smooth objective function with proved convergence. Empirical studies on cognitive data of the ADNI cohort demonstrate the superior performance of the proposed method. 

Summary It is well‐known that the mycorrhizal type of plants correlates with different modes of nutrient cycling and availability. However, the differences in drought tolerance between arbuscular mycorrhizal (AM) and ectomycorrhizal (EcM) plants remains poorly characterized.We synthesized a global dataset of four hydraulic traits associated with drought tolerance of 1457 woody species (1139 AM and 318 EcM species) at 308 field sites. We compared these traits between AM and EcM species, with evolutionary history (i.e. angiosperms vs gymnosperms), water availability (i.e. aridity index) and biomes considered as additional factors.Overall, we found that evolutionary history and biogeography influenced differences in hydraulic traits between mycorrhizal types. Specifically, we found that (1) AM angiosperms are less drought‐tolerant than EcM angiosperms in wet regions or biomes, but AM gymnosperms are more drought‐tolerant than EcM gymnosperms in dry regions or biomes, and (2) in both angiosperms and gymnosperms, variation in hydraulic traits as well as their sensitivity to water availability were higher in AM species than in EcM species.Our results suggest that global shifts in water availability (especially drought) may alter the biogeographic distribution and abundance of AM and EcM plants, with consequences for ecosystem element cycling and ultimately, the land carbon sink.