蚁群优化(ACO)算法是一种常用的元启发式算法,它通过模拟蚁群寻找食物的过程,为求解多维背包问题(MKP)等NP难(Non-deterministic Polynomial hard)问题提供可行途径。原始ACO算法及其改进算法,通常分为多个轮次,每个轮次均会生成一个蚂...蚁群优化(ACO)算法是一种常用的元启发式算法,它通过模拟蚁群寻找食物的过程,为求解多维背包问题(MKP)等NP难(Non-deterministic Polynomial hard)问题提供可行途径。原始ACO算法及其改进算法,通常分为多个轮次,每个轮次均会生成一个蚂蚁种群寻找可行解。在不同轮次中,每轮蚁群中蚂蚁的数量是固定的,因此,如果将其指定一个较大的值,会导致算法出现不必要的时间消耗;反之,如果指定的值较小,则会降低算法全局最优解搜索能力。为此,提出了一种基于蚁群数量动态调整的改进蚁群优化算法ACO-ANDA(ACO algorithm based on Ant Number Dynamic Adjustment),所提算法在可行解搜索过程中,引入了一种新的蚁群数量动态调整机制。在每轮可行解搜索结束后,均根据近几轮可行解和历史最优解之间的关系,调整下一轮蚁群数量,实现对算法时间耗费和最优解搜索能力的平衡。再基于MKP基准测试集SAC-94的多组实验结果表明,相较于原始ACO算法,所提算法能够在最优解利润平均降低0.02%的情况下,平均降低77.85%的时间耗费。展开更多
Patients treated with the cisplatin often develop strong resistance to the drug after prolonged treatments, ultimately resulting in limited clinical efficacy. One of the possible mechanisms is that the internalized co...Patients treated with the cisplatin often develop strong resistance to the drug after prolonged treatments, ultimately resulting in limited clinical efficacy. One of the possible mechanisms is that the internalized compound may be inactivated before getting access to the nucleus where cisplatin forms a complex with the genomic DNA and triggers a cell death program. However, the nature and intracellular fate of inactivated cisplatin is poorly illustrated. In the present study, we reported for the first time the presence of platinum nanoparticles(Pt-NPs) in the cytoplasm of cells treated with cisplatin. Further analysis also evidenced a correlation of the increased intracellular PtNPs formation with cisplatin resistance, and confirmed the process was glutathione S-transferase relevant. Our data suggest that tumor cells may develop cisplatin resistance by converting the drug into less toxic intracellular Pt-NPs, thereby impeding the drug from targeting its substrates.展开更多
文摘蚁群优化(ACO)算法是一种常用的元启发式算法,它通过模拟蚁群寻找食物的过程,为求解多维背包问题(MKP)等NP难(Non-deterministic Polynomial hard)问题提供可行途径。原始ACO算法及其改进算法,通常分为多个轮次,每个轮次均会生成一个蚂蚁种群寻找可行解。在不同轮次中,每轮蚁群中蚂蚁的数量是固定的,因此,如果将其指定一个较大的值,会导致算法出现不必要的时间消耗;反之,如果指定的值较小,则会降低算法全局最优解搜索能力。为此,提出了一种基于蚁群数量动态调整的改进蚁群优化算法ACO-ANDA(ACO algorithm based on Ant Number Dynamic Adjustment),所提算法在可行解搜索过程中,引入了一种新的蚁群数量动态调整机制。在每轮可行解搜索结束后,均根据近几轮可行解和历史最优解之间的关系,调整下一轮蚁群数量,实现对算法时间耗费和最优解搜索能力的平衡。再基于MKP基准测试集SAC-94的多组实验结果表明,相较于原始ACO算法,所提算法能够在最优解利润平均降低0.02%的情况下,平均降低77.85%的时间耗费。
文摘金融时间序列预测是经济领域中一个非常重要的实际问题,然而,由于金融市场的噪声和波动性,当前存在方法的预测精度尚不能令人满意。为了提高金融时间序列的预测精度,提出了一种融合扩张卷积神经网络(dilated convolutional neural network,DCNN)、长短时记忆神经网络(long short term memory,LSTM)和注意力机制(attention mechanism,AT)的混合预测模型DCNN_LSTM_AT。该模型由两个部分组成:第一部分包含扩张卷积神经网络和基于LSTM的编码器,其功能在于提取原始序列数据中不同时间尺度的有效信息;第二部分由带注意力机制的LSTM解码器构成,其功能在于对第一部分提取的信息进行过滤并利用过滤后的信息进行预测。最后将所提模型在3支股指数据集和3支个股数据集上进行实验,并与其他常见的基准模型进行了对比,实验结果表明该模型相比于其他模型具有更好的预测精度和稳定性.
基金Funded by the National Key Basic Research Program of China (2011CB933503)the National Natural Science Foundation of China (NSFC) for Key Project of International Cooperation (61420106012)+2 种基金the Special Funds of National Natural Science Foundation of China For Basic Research Projects of Scientific Instruments (61127002)the Special Project on the Development of National Key Scientific Instruments and Equipment of China (2011YQ03013403)China Postdoctoral Science Foundation funded project (2013M541592)
文摘Patients treated with the cisplatin often develop strong resistance to the drug after prolonged treatments, ultimately resulting in limited clinical efficacy. One of the possible mechanisms is that the internalized compound may be inactivated before getting access to the nucleus where cisplatin forms a complex with the genomic DNA and triggers a cell death program. However, the nature and intracellular fate of inactivated cisplatin is poorly illustrated. In the present study, we reported for the first time the presence of platinum nanoparticles(Pt-NPs) in the cytoplasm of cells treated with cisplatin. Further analysis also evidenced a correlation of the increased intracellular PtNPs formation with cisplatin resistance, and confirmed the process was glutathione S-transferase relevant. Our data suggest that tumor cells may develop cisplatin resistance by converting the drug into less toxic intracellular Pt-NPs, thereby impeding the drug from targeting its substrates.
