期刊文献+
共找到2篇文章
< 1 >
每页显示 20 50 100
基于RF的排序学习方法在电影数据集中的应用 被引量:1
1
作者 何启泓 李旭军 孙燕 《电脑与信息技术》 2021年第5期1-6,共6页
针对自制电影数据集中电影的排序问题,文章提出了一种基于RF的Bootstrap自适应双集成排序学习方法(RandomForest-based Bootstrap Self-adaptive Double-ensemble,RF-based BSD)。先利用电影媒体网站数据构建21个特征自建基于排序学习... 针对自制电影数据集中电影的排序问题,文章提出了一种基于RF的Bootstrap自适应双集成排序学习方法(RandomForest-based Bootstrap Self-adaptive Double-ensemble,RF-based BSD)。先利用电影媒体网站数据构建21个特征自建基于排序学习格式的电影数据集,BSD会根据输入数据集的查询数、查询-电影对数和特征数,通过Bootstrap自适应函数自动确定RF的子采样比例,然后使用单集成模型(比如MART,Multiple Additive Regression Tree,多重累计回归树)作为基学习器进行训练,最后采用bagging思想输出最终的双集成模型。实验结果显示,对比两个评价指标NDCG(Normalized Discounted Cumulative Gain,归一化折扣累计增益)和MAP(Mean Average Precision,平均值均值)的评估效果,发现BSD输出的双集成模型比单集成模型在两项指标上均有1%-3%左右的提升。 展开更多
关键词 随机森林 排序学习 电影数据集 Bootstrap子采样 成模型
下载PDF
Surface roughness modulated resistivity in copper thin films
2
作者 HongKang Song Ke Xia Jiang Xiao 《Science China(Physics,Mechanics & Astronomy)》 SCIE EI CAS CSCD 2018年第10期70-77,共8页
The surfaces of metallic thin films are never flat. The resistivity in thin films is very different from that in bulk because of the unavoidable rough surfaces. In this study, we apply a quantum-mechanical method to s... The surfaces of metallic thin films are never flat. The resistivity in thin films is very different from that in bulk because of the unavoidable rough surfaces. In this study, we apply a quantum-mechanical method to study the resistivity in metallic thin films.The resulting resistivity formula for metallic thin films merely involves two parameters: bulk relaxation time and surface roughness. We use the formula to fit a large number of experimental data sets for copper thin films obtained using different growing methods. With an additional tuning parameter for calibrating the film thickness, the quantum formula can provide a universal fitting to most data with a satisfactory precision, regardless of their growing methods or data source. 展开更多
关键词 thin film thickness dependent RESISTIVITY surface roughness
原文传递
上一页 1 下一页 到第
使用帮助 返回顶部