Search

Cost-sensitive learning is a popular approach to addressing the problem of class imbalance for many classification algorithms in machine learning. However, most cost-sensitive algorithms are dependent on manually designed cost matrices. Unfortunately, in many cases, it is often not easy for humans, even experts, to accurately specify misclassification costs for different mistakes due to the lack of domain knowledge related to actual situations in some complex unbalanced problems. As a result, these cost-sensitive algorithms cannot be directly applied. This paper proposes a new genetic programmingbased approach to developing cost-sensitive classifiers that are independent of manually designed cost matrices. The proposed method is able to construct classifiers and learn cost intervals automatically and simultaneously. In the proposed method, a tree representation, terminal sets and function sets are designed and developed. We examine the effectiveness of the proposed method on ten high-dimensional unbalanced datasets. The experimental results show that the proposed method often outperforms compared methods for highdimensional unbalanced classification. Furthermore, according to the analysis of evolved trees, the constructed classifiers often only need a small number of features to achieve a good classification performance.

Tumor microenvironment (TME) has been shown to be extensively involved in tumor development. However, the dynamic change of TME components and their effects are still unclear. Here, we attempted to identify TME-related genes that could help predict survival and may be potential therapeutic targets.Data was collected from UCSC Xena and GEO database. ESTIMATE and CIBERSORT algorithms were applied to estimate the components and the proportions of TIICs in TME. We analyzed the gene expression differences of immune components and stromal components, respectively, and finally got the overlapped DEGs. Through protein-protein interaction (PPI) network and univariate Cox regression analysis based on shared DEGs, we screened out and validated the TME-related genes. Focusing on this gene, we analyzed the expression and prognostic value of this gene, and investigated its relationship with immune cells by correlation analysis, single cell analysis, immunohistochemistry and immunofluorescence analysis.Through a series analysis, we found that the proportion of immune and stromal components was an important prognostic factor, and screened out a key gene, LPAR5, which was highly correlated with prognosis and metastasis. And the expression of LPAR5 was positively correlated with immune cells, especially macrophages, indicating LPAR5We identified a TME-related gene, LPAR5, which is a promising indicator for TME remodeling in osteosarcoma. Particularly, LPAR5

Osteoarthritis (OA), characterized by chronic systemic low-level inflammation and cartilage degeneration, is a type of arthritis closely associated with aging. Inflammation and aging play a pivotal role in the occurrence and progression of OA. NLRP3 inflammasome is involved in many inflammatory and aging diseases, and NLRP3 inhibitor MCC950 has anti-inflammatory and antisenescence effects on some diseases such as Alzheimer’s disease. In the present study, we found that NLRP3 protein was upregulated in human and mouse OA cartilage. Moreover, NLRP3 and Caspase1 expression induced by IL-1β in chondrocytes was blocked by MCC950. In addition, MCC950 inhibited the expression of inflammatory mediators, matrix-degrading enzymes, senescence marker protein P16 (INK4A), and β-galactosidase, as well as excessive production of ROS. Meanwhile, MCC950 promoted autophagy-related protein expression and autophagy flux under the inflammatory condition. However, autophagy inhibitor 3-MA reversed anti-inflammatory and anticatabolic effects of MCC950. In in vivo experiments, intra-articular administration of MCC950 further showed its protective effect on cartilage degeneration. Bioinformatic analysis and in vitro experimental results revealed that MCC950 might play a protective role in cartilage by regulating Nrf2/HO-1/NQO1, PI3k/Akt/mTOR, P38/MAPK, and JNK/MAPK pathways. In conclusion, our work demonstrated that NLRP3 inhibitor MCC950 might serve as a promising strategy for OA treatment.

For deep convolutional neural networks (deep CNNs), a severe drawback is the poor interpretability. To address this drawback, this paper proposes a novel genetic algorithm-based method for the first time to automatically evolve local interpretable explanations that can assist users to decide whether to trust the predictions of deep CNNs. In the experiments, the results show that the evolved explanations can explain the predictions of deep CNNs on images by successfully capturing meaningful interpretable features.

In classification, when class overlap is intertwined with the issue of class imbalance, it is often challenging to discover useful patterns because of an ambiguous boundary between the majority class and the minority class. This becomes more difficult if the data is high-dimensional. To date, very few pieces of work have investigated how the class overlap issue can be effectively addressed or alleviated in classification with high-dimensional unbalanced data. In this paper, we propose a new genetic programming based method, which is able to automatically and directly detect borderline instances, in order to address the class overlap issue in classification with high-dimensional unbalanced data. In the proposed method, each individual has two trees to be trained together based on different classification rules. The proposed method is examined and compared with baseline methods on high-dimensional unbalanced datasets. Experimental results show that the proposed method achieves better classification performance than the baseline methods in almost all cases.

Osteoarthritis (OA) is a common degenerative joint disease blamed for pain and disability in the elderly. Galangin (GAL) is a natural flavonoid that exhibits anti-inflammatory properties in various inflammation diseases. However, the role of GAL in OA remains unclear. In this study, we investigate the role of GAL in the progress and development of OA in vitro and vivo. The results showed that IL-1β exposure resulted in increased expression of iNOS, COX-2, MMP1, MMP3, MMP13 and ADAMTS5 in rat chondrocytes. However, co-treatment with GAL significantly decreased theses inflammatory cytokines and catabolic factors expression. In addition, GAL reduced IL-1β-induced degradation of collagen II and aggrecan in chondrocytes. Furthermore, GAL significantly suppressed IL-1β-induced Akt phosphorylation and NF-κB activation in rat chondrocytes. In vivo, intra-articular injection of GAL could also reduce the cartilage degradation in the ACLT rat model. This study reveals galangin may act as a promising novel agent in the treatment of OA.

The sound velocities of water in the Hugoniot states are investigated by laser shock compression of precompressed water in a diamond anvil cell. The obtained sound velocities in the off-Hugoniot region of liquid water at precompressed conditions are used to test the predictions of quantum molecular dynamics (QMD) simulations and the SESAME equation-of-state (EOS) library. It is found that the prediction of QMD simulations agrees with the experimental data while the prediction of SESAME EOS library underestimates the sound velocities probably due to its improper accounting for the ionization processes.

High-dimensional unbalanced classification is challenging because of the joint effects of high dimensionality and class imbalance. Genetic programming (GP) has the potential benefits for use in high-dimensional classification due to its built-in capability to select informative features. However, once data are not evenly distributed, GP tends to develop biased classifiers which achieve a high accuracy on the majority class but a low accuracy on the minority class. Unfortunately, the minority class is often at least as important as the majority class. It is of importance to investigate how GP can be effectively utilized for high-dimensional unbalanced classification. In this article, to address the performance bias issue of GP, a new two-criterion fitness function is developed, which considers two criteria, that is, the approximation of area under the curve (AUC) and the classification clarity (i.e., how well a program can separate two classes). The obtained values on the two criteria are combined in pairs, instead of summing them together. Furthermore, this article designs a three-criterion tournament selection to effectively identify and select good programs to be used by genetic operators for generating offspring during the evolutionary learning process. The experimental results show that the proposed method achieves better classification performance than other compared methods.

Many real-world classification problems encounter the problem of class imbalance. Genetic programming (GP) is one of the most important evolutionary algorithms, making crucial contributions to classification, particularly in the high-dimensional case. However, like other classification algorithms, GP may also develop biased classifiers when the class distribution is unbalanced or skewed. This is because standard GP treats each instance equally and assumes the same cost of different misclassification. In unbalanced classification, misclassification cost of the minority class is often serious than that of the majority class. Cost-sensitive learning has been successfully applied to solve the problem of class imbalance for many classification algorithms, but it has not been heavily investigated in GP. This paper investigates how cost-sensitive learning can be effectively used by GP to address the problem of class imbalance in high-dimensional unbalanced classification. Experimental results on six high-dimensional unbalanced datasets show the better performance of the proposed methods than the compared methods.

High-dimensionality and class imbalance represent two main challenges in classification. Recently, there is a growing number of datasets exhibiting the characteristics of the combination of the class imbalance and high-dimensionality. Genetic programming (GP) has been successfully applied to solve high-dimensional classification tasks. However, most existing GP methods may also suffer from a performance bias if the class distribution is unbalanced. Using fitness functions for cost adjustment is one of the most important methods in GP to address the class imbalance issue. This paper develops new fitness functions in GP to address the class imbalance issue in classification with high-dimensional unbalanced data. Two fitness functions are proposed to increase the performance of the traditional accuracy measures, and one fitness function is proposed to approximate Area Under Curve (AUC) with the goal to save the training time. Experiments on six high-dimensional unbalanced datasets show the better performance of the proposed fitness functions, compared to existing fitness functions.

Genetic programming (GP) has the built-in ability for feature selection when developing classifiers for classification with high-dimensional data. However, due to the problem of class imbalance, the developed classifiers by GP are prone to be biased towards the majority class. Cost-sensitive learning has shown to be effective in addressing the problem of class imbalance. In cost-sensitive learning, cost matrices are often manually designed and then considered by classification algorithms to treat different mistakes differently. However, in many real-world applications, cost matrices are unknown because of the limited domain knowledge in complex situations. Therefore, in this paper, we propose a novel GP method to develop cost-sensitive classifiers, where a cost matrix is automatically learned, instead of requiring it from domain experts. The proposed method is examined and compared with existing methods on ten high-dimensional unbalanced datasets. Experimental results show that the proposed method outperforms the compared GP methods in most cases.

Classification tasks become more challenging when having the curse of dimensionality issue. Recently, there has been an increasing number of datasets with thousands of features. Some classification algorithms often need feature selection to avoid the curse of dimensionality. Genetic programming (GP) has shown success in classification tasks. GP does not require to do feature selection because of its built-in capability to automatically select informative features. However, GP-based methods are often computationally intensive to achieve a good classification accuracy. Based on perspectives from granular computing (GrC), this paper proposes a new approach to linking features hierarchically for GP-based classification. Experiments on seven high-dimensional datasets show the effectiveness of the proposed algorithm in terms of saving training time and enhancing the classification accuracy, compared to baseline methods.

This paper focuses on the sorting problems with multiple types of attributes. About the attributes, in which are divided into qualitative attributes, quantitative attributes, qualitative criteria and quantitative criteria. Granules of knowledge are defined by applying four types of relations simultaneously: indiscernibility relation defined on qualitative attributes, similarity relation defined on quantitative attributes, dominance relation defined on qualitative criteria and quasi-partial order relation defined on quantitative criteria. To guarantee the tolerance of the system, the threshold is adjusted, resulting in a N-neighborhood system comes into being. The consistency measure which possess properties of monotonicity is regarded as the Likelihood Function, so the optimal threshold is obtained by Maximum Likelihood Estimation, as a result, N-neighborhood system is converted into optimal 1-neighborhood system. Therefore, we proposed the Optimal-Neighborhood Statistics Rough Set Approach with Multiple Attributes and Criteria.

A technique that combines the diagnostics of electron temperature history and the measurements of ablation velocity with two-tracer x-ray spectroscopy has been developed for diagnosing the temperature profiles in the thermal conduction layers of laser-ablated plastic foils. The electron temperature in the plastic ablator was diagnosed using the isoelectronic line ratios of Al Lyα line to Mg Lyα line, emitted from a tracer layer of Al/Mg mixture buried under the ablator. The ablation velocity was inferred from the time delay between the onset time of x-ray line emissions from Al and Mg tracer layers buried at two depths in the ablator, respectively. From the measured electron temperatures and ablation velocity, the electron temperature profile in the conduction layer was inferred. The measured temperature profile was compared with the simulated one and reasonable agreement was found.