Tubal-Sampling: Bridging Tensor and Matrix Completion in 3-D Seismic Data Reconstruction

Abstract

Forecasting project expenses is a crucial step for businesses to avoid budget overruns and project failures. Traditionally, this has been done by financial analysts or data science techniques such as time-series analysis. However, these approaches can be uncertain and produce results that differ from the planned budget, especially at the start of a project with limited data points. This paper proposes a constrained non-negative matrix completion model that predicts expenses by learning the likelihood of the project correlating with certain expense patterns in the latent space. The model is constrained on three probability simplexes, two of which are on the factor matrices and the third on the missing entries. Additionally, the predicted expense values are guaranteed to meet the budget constraint without the need of post-processing. An inexact alternating optimization algorithm is developed to solve the associated optimization problem and is proven to converge to a stationary point. Results from two real datasets demonstrate the effectiveness of the proposed method in comparison to state-of-the-art algorithms.

3-D Seismic Noise Attenuation via Tensor Sparse Coding With Spatially Adaptive Coherence Constraint

“…Definition 4 (f-Diagonal Tensor [38]): A tensor is called f-diagonal as each frontal slice of the tensor is a diagonal matrix. [38], [39]): The t-SVD of Y is defined as (see Fig. 1)…”

Robust fast dictionary learning for seismic noise attenuation

“…To date, researchers have proposed many noise suppression methods, including filtering‐based approaches (Abma and Claerbout, 1995; Wang, 1999; Gan et al ., 2016), decomposition‐based approaches (Huang et al ., 1998; Bekara and Van der Baan, 2007; Fomel, 2013; Forghani‐Arani et al ., 2013; Han and van der Baan, 2015; Gómez and Velis, 2016), compressed‐sensing‐based methods (Fomel and Liu, 2010; Mousavi et al ., 2016; Iqbal et al ., 2016; Tang et al ., 2018; Li et al ., 2018; Shao et al ., 2019) and deep‐learning‐based methods (Saad and Chen, 2020b; Qian et al ., 2020, 2021; Saad et al ., 2021; Yang et al ., 2022). The filtering‐based method, such as median‐filtering‐based and predictive‐filtering‐based approaches, is implemented by designing a filter for the seismic data and eliminating the unwanted seismic data components.…”

A Comment Aspect-Level User Preference Transfer Model for Cross-Domain Recommendations

“…User data from different fields is not always random; for example, users who enjoy detective books may like detective movies (Sang et al, 2023;Huang et al, 2023). This potentially similar pattern in different fields makes cross-domain recommendation (CDR) possible to meet the needs of user preference migration between different fields in practical applications (Qian et al, 2020;Soydaner, 2022). If users' preferences in one domain can be transferred to another domain, it will help improve the accuracy and efficiency of recommendations.…”