Visual domain adaptation aims to learn robust classi ers for the target domain by leveraging knowledge from a source domain. Existing methods either a empt to align the cross-domain distributions, or perform manifold subspace learning. However, there are two signi cant challenges: (1) degenerated feature transformation, which means that distribution alignment is o en performed in the original feature space, where feature distortions are hard to overcome. On the other hand, subspace learning is not su cient to reduce the distribution divergence. (2) unevaluated distribution alignment, which means that existing distribution alignment methods only align the marginal and conditional distributions with equal importance, while they fail to evaluate the di erent importance of these two distributions in real applications. In this paper, we propose a Manifold Embedded Distribution Alignment (MEDA) approach to address these challenges. MEDA learns a domain-invariant classi er in Grassmann manifold with structural risk minimization, while performing dynamic distribution alignment to quantitatively account for the relative importance of marginal and conditional distributions. To the best of our knowledge, MEDA is the rst a empt to perform dynamic distribution alignment for manifold domain adaptation. Extensive experiments demonstrate that MEDA shows signi cant improvements in classi cation accuracy compared to state-of-the-art traditional and deep methods. * e rst two authors contributed equally. † J. Wang and Y. Chen are also a liated with Beijing Key Lab. of Mobile Computing and Pervasive Devices. W. Feng is also with CAS Key Lab. of Network Data Science & Technology. J. Wang and W. Feng are also a liated with University of Chinese Academy of Sciences. ‡ P. Yu is also a liated
Transfer learning has achieved promising results by leveraging knowledge from the source domain to annotate the target domain which has few or none labels. Existing methods often seek to minimize the distribution divergence between domains, such as the marginal distribution, the conditional distribution or both. However, these two distances are often treated equally in existing algorithms, which will result in poor performance in real applications. Moreover, existing methods usually assume that the dataset is balanced, which also limits their performances on imbalanced tasks that are quite common in real problems. To tackle the distribution adaptation problem, in this paper, we propose a novel transfer learning approach, named as Balanced Distribution Adaptation (BDA), which can adaptively leverage the importance of the marginal and conditional distribution discrepancies, and several existing methods can be treated as special cases of BDA. Based on BDA, we also propose a novel Weighted Balanced Distribution Adaptation (W-BDA) algorithm to tackle the class imbalance issue in transfer learning. W-BDA not only considers the distribution adaptation between domains but also adaptively changes the weight of each class. To evaluate the proposed methods, we conduct extensive experiments on several transfer learning tasks, which demonstrate the effectiveness of our proposed algorithms over several state-of-the-art methods.
Transfer learning aims to learn robust classifiers for the target domain by leveraging knowledge from a source domain. Since the source and the target domains are usually from different distributions, existing methods mainly focus on adapting the cross-domain marginal or conditional distributions. However, in real applications, the marginal and conditional distributions usually have different contributions to the domain discrepancy. Existing methods fail to quantitatively evaluate the different importance of these two distributions, which will result in unsatisfactory transfer performance. In this paper, we propose a novel concept called Dynamic Distribution Adaptation (DDA), which is capable of quantitatively evaluating the relative importance of each distribution. DDA can be easily incorporated into the framework of structural risk minimization to solve transfer learning problems. On the basis of DDA, we propose two novel learning algorithms: (1) Manifold Dynamic Distribution Adaptation (MDDA) for traditional transfer learning, and (2) Dynamic Distribution Adaptation Network (DDAN) for deep transfer learning. Extensive experiments demonstrate that MDDA and DDAN significantly improve the transfer learning performance and setup a strong baseline over the latest deep and adversarial methods on digits recognition, sentiment analysis, and image classification. More importantly, it is shown that marginal and conditional distributions have different contributions to the domain divergence, and our DDA is able to provide good quantitative evaluation of their relative importance which leads to better performance. We believe this observation can be helpful for future research in transfer learning.
Background/Aims: Next-generation sequencing (NGS) has revealed abundant long noncoding RNAs (lncRNAs) that have been characterized as critical components of cancer biology in humans. The present study aims to investigate the role of the lncRNA KCNQ1OT1 in breast cancer (BRCA) as well as the underlying molecular mechanisms and functions of KCNQ1OT1 involved in the progression of BRCA. Methods: The Cancer Genome Atlas (TCGA) and StarBase v2.0 were used to obtain the required gene data. Dual luciferase reporter gene assays were conducted to verify the relevant intermolecular target relationships. QRT-PCR and Western blot were performed to measure the expression levels of different molecules. Cell proliferation was detected by using the MTT and colony formation assays, while cell migration and invasion were examined by transwell assay. Variations in cell apoptosis and cell cycle were determined through flow cytometry. A tumor xenograft model was applied to assess tumor growth in vivo. Results: KCNQ1OT1 was found to be remarkably highly expressed in BRCA tissues and cells. KCNQ1OT1 modulated CCNE2 through sponging miR-145 in BRCA. KCNQ1OT1 promoted tumor growth in vivo by regulating miR-145/CCNE2. Conclusion: The KCNQ1OT1/miR-145/CCNE2 axis plays a critical regulatory role in BRCA, potentially giving rise to BRCA tumorigenesis and progression. These findings provide valuable evidence for improving the diagnosis and treatment of BRCA in the future.
These findings suggest that ZFX plays an important role in breast cancer development and could be a potential therapeutic target for breast cancer.
Abstract. Paired box 6 (PAX6) plays a significant role in the development of human neuroectodermal epithelial tissues. Previous studies have suggested that the PAX6 promoter is hypermethylated in breast cancer and that it is involved in breast cancer cell proliferation. The present study aimed to investigate the expression of PAX6 in invasive breast cancer tissues, and to evaluate its prognostic significance. Immunohistochemistry (IHC) was used to detect PAX6 expression on a breast cancer tissue microarray containing tissues from 111 patients. Associations of PAX6 expression with staging and prognosis were analyzed. PAX6 was mainly expressed in the nucleus. The PAX6 staining intensity was not associated with age, histological grade, lymph node status, tumor size, or progesterone receptor and human epidermal growth factor receptor 2 expression (all P>0.05). A high level of PAX6 staining was more frequent in estrogen receptor (ER)-negative cases compared with ER-positive cases (43.9 vs. 25.7%; P= 0.049). After a median follow-up time of 110 months, the patients with low PAX6 expression exhibited an improved survival rate compared with the patients with high PAX6 expression (P<0.001). Cox analysis showed a worse survival rate in the patients with high PAX6 staining (hazard ratio, 3.458; 95% confidence interval, 1.575-7.593; P= 0.002). In conclusion, high tumor PAX6 staining intensity by IHC was associated with a poor prognosis in breast cancer patients.
Asparagine synthetase (ASNS) is deemed to be a promising therapeutic target for the treatment of several cancers, but its functional role in human breast cancer is still unknown. In this study, we employed RNA interference as an efficient tool to silence endogenous ASNS expression in breast cancer cell lines. The relationship between ASNS expression and breast cancer cell growth was investigated, and the therapeutic value of ASNS in breast cancer was further evaluated. Depletion of ASNS remarkably inhibited the proliferation and colony formation capacity of breast cancer cells and arrested cell cycle in the S phase. Our findings suggest that ASNS may contribute to breast cancer tumorigenesis and could be a potential therapeutic target in human breast cancer.
Despite the knowledge of many genetic alterations present in breast cancer, the complexity of this disease precludes placing its biology into a simple conceptual framework. Toll-like receptor 4 (TLR4) plays important roles in regulating innate immunity and may affect the development of cancers. Polymorphisms in TLR4 gene have been shown to be associated with impaired immune responses. Here, we investigated the association of TLR4 polymorphisms with breast cancer. Four functional TLR4 polymorphisms (-2242T/C, Asp299Gly, Thr399Ile, and +3725G/C) were genotyped in a total of 665 breast cancer patients and 768 healthy controls. Data were analyzed using the chi-squared test. Results showed that the prevalence of TLR4 +3725GC and CC genotypes were significantly increased in breast cancer cases when compared with controls [odds ratio (OR) = 1.37, 95% confidence interval (CI) = 1.08-1.73, P = 0.008 and OR = 2.34, 95% CI = 1.66-3.35, P < 0.0001, respectively]. Also, the frequency of TLR4 +3725C allele was significantly higher in breast cancer patients (P < 0.0001). The -2242T/C polymorphism did not show any significant differences between cases and controls. In addition, when analyzing the survival time of breast cancer patients with TLR4 +3725G/C polymorphism, cases with +3725C allele had significantly shorter survival time overall (P = 0.006). These results suggested that polymorphism in TLR4 gene was associated with increased susceptibility to breast cancer and could be used as a prognostic marker for this malignancy.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.
hi@scite.ai
10624 S. Eastern Ave., Ste. A-614
Henderson, NV 89052, USA
Copyright © 2024 scite LLC. All rights reserved.
Made with 💙 for researchers
Part of the Research Solutions Family.