This paper presents a novel approach called Graph Embedded Stochastic Configuration Networks (GESCNs) for complex data classification. To enhance the learning of geometric and discriminative features, especially in imbalanced and noisy datasets, the proposed GESCNs method integrates graph embedding strategies with Stochastic Configuration Networks (SCNs), a novel incremental randomized learning method known for its efficient model configuration mechanism. The intrinsic and penalty graph matrices help preserve topological relationships between data points, while SCNs ensure effective learning and strong generalization in modeling. Extensive experiments on diverse datasets demonstrate that GESCNs consistently outperform other randomized learning models across key metrics. These findings underscore the robustness of GESCNs in handling noisy and imbalanced data, making it a highly effective solution for challenging classification tasks.

Seismic activity prediction is usually considered an infeasible and impossible task due to its intricate underlying mechanism. However, the rapid development of high-performance monitoring systems and artificial intelligence techniques for complex data modeling has reignited interest in exploring failure precursors and predicting seismic events. In this paper, we employ Stochastic Configuration Networks for problem-solving. To explore laboratory seismic activity deeply, we apply SCN to learn essential seismic information for the time-to-failure (i.e., time remaining before the occurrence of seism) prediction with acoustic emission signals. To alleviate overfitting during the learning process, a regularized SCN is used. Interestingly, we replace the standard Sigmoid activation function with the Hard Sigmoid, which accelerates convergence and improves performance further. The experimental results demonstrate that the improved method outperforms other machine learning methods regarding coefficient of determination and Mean Absolute Error.

Speech-based automatic smoker identification (also known as smoker/non-smoker classification) aims to identify speakers’ smoking status from their speech. In the COVID-19 pandemic, speech-based automatic smoker identification approaches have received more attention in smoking cessation research due to low cost and contactless sample collection. This study focuses on determining the best acoustic features for smoker identification. In this paper, we investigate the performance of four acoustic feature sets/representations extracted using three feature extraction/learning approaches: (i) hand-crafted feature sets including the extended Geneva Minimalistic Acoustic Parameter Set and the Computational Paralinguistics Challenge Set, (ii) the Bag-of-Audio-Words representations, (iii) the neural representations extracted from raw waveform signals by SincNet. Experimental results show that: (i) SincNet feature representations are the most effective for smoker identification and outperform the MFCC baseline features by 16% in absolute accuracy; (ii) the performance of hand-crafted feature sets and the Bag-of-Audio-Words representations rely on the scale of the dimensions of feature vectors.

We propose a novel pitch estimation technique called DeepF0, which leverages the available annotated data to directly learns from the raw audio in a data-driven manner. f0estimation is important in various speech processing and music information retrieval applications. Existing deep learning models for pitch estimations have relatively limited learning capabilities due to their shallow receptive field. The proposed model addresses this issue by extending the receptive field of a network by introducing the dilated convolutional blocks into the network. The dilation factor increases the network receptive field exponentially without increasing the parameters of the model exponentially. To make the training process more efficient and faster, DeepF0 is augmented with residual blocks with residual connections. Our empirical evaluation demonstrates that the proposed model outperforms the baselines in terms of raw pitch accuracy and raw chroma accuracy even using 77.4% fewer network parameters. We also show that our model can capture reasonably well pitch estimation even under the various levels of accompaniment noise.