Hyperspectral band selection plays a pivotal role in remote sensing and image analysis, aiming to identify the most informative spectral bands while minimizing computational overhead. In this paper, we introduce a pioneering approach for hyperspectral band selection that offers an embedded solution, making it well-suited for resource-constrained or real-time applications. Our proposed method, embedded Hyperspectral Band Selection (EHBS), excels in selecting the best bands without the need for prior processing, seamlessly integrating with the downstream task model. This is achieved through the adaptation of the Stochastic Gates (STG) algorithm, originally designed for feature selection, for hyperspectral band selection in the context of image semantic segmentation and the integration of a dynamic optimizer, DoG, which removes the need for the required tuning the learning rate. To assess the performance of our method, we introduce a novel metric for evaluating band selection methods across different target numbers of selected bands quantified by the Area Under the Curve (AUC). We conduct experiments on two distinct semantic-segmentation hyperspectral benchmark datasets, demonstrating its superiority in terms of its resulting accuracy and its ease of use compared to many common and state-of-the-art methods. Furthermore, our contributions extend beyond the realm of hyperspectral band selection. The adaptability of our approach to other tasks, especially those involving grouped features, opens up promising avenues for broader applications within the realm of deep learning, such as feature selection for feature groups. The demonstrated success on the tested datasets and the potential for application to a variety of tasks underscore the value of our method as a substantial addition to the field of computer vision.
In the realm of machine and deep learning regression tasks, the role of effective feature engineering (FE) is pivotal in enhancing model performance. Traditional approaches of FE often rely on domain expertise to manually design features for machine learning models. In the context of deep learning models, the FE is embedded in the neural network's architecture, making it hard for interpretation. In this study, we propose to integrate symbolic regression (SR) as an FE process before a machine learning model to improve its performance. We show, through extensive experimentation on synthetic and real-world physics-related datasets, that the incorporation of SR-derived features significantly enhances the predictive capabilities of both machine and deep learning regression models with 34-86% root mean square error (RMSE) improvement in synthetic datasets and 4-11.5% improvement in real-world datasets. In addition, as a realistic use-case, we show the proposed method improves the machine learning performance in predicting superconducting critical temperatures based on Eliashberg theory by more than 20% in terms of RMSE. These results outline the potential of SR as an FE component in data-driven models.
This paper studies the potential of identifying lexical paraphrases within a single corpus, focusing on the extraction of verb paraphrases. Most previous approaches detect individual paraphrase instances within a pair (or set) of comparable corpora, each of them containing roughly the same information, and rely on the substantial level of correspondence of such corpora. We present a novel method that successfully detects isolated paraphrase instances within a single corpus without relying on any a-priori structure and information. A comparison suggests that an instance-based approach may be combined with a vector based approach in order to assess better the paraphrase likelihood for many verb pairs.