HIGGS Signal Events for Machine Learning

Signals (y=1) events from the first one million events in the HIGGS machine learning dataset at: http://archive.ics.uci.edu/ml/datasets/HIGGS The data has been produced using Monte Carlo simulations. The first 21 features (columns 2-22) are kinematic properties measured by the particle detectors in the accelerator. The last seven features are functions of the first 21 features; these are high-level features derived by physicists to help discriminate between the two classes. There is an interest in using deep learning methods to obviate the need for physicists to manually develop such features. Benchmark results using Bayesian Decision Trees from a standard physics package and 5-layer neural networks are presented in a paper here: http://rdcu.be/cb58. In the Nature paper the full dataset was used, not just hte first 1million which were uploaded here.

本数据集选取了HIGGS机器学习数据集（HIGGS machine learning dataset）前一百万条事件中的标记为y=1的信号事件，原始数据集的公开地址为：http://archive.ics.uci.edu/ml/datasets/HIGGS。所有数据均通过蒙特卡洛模拟（Monte Carlo simulations）生成。前21个特征（对应第2至22列）为加速器粒子探测器所测得的运动学特性。剩余7个特征为前21个特征的函数组合，此类特征为物理学家手动推导得到的高阶特征，用于区分两类样本。当前学界存在利用深度学习方法替代物理学家手动构造此类特征的研究诉求。一篇《自然》（Nature）期刊论文中给出了采用标准物理工具包中的贝叶斯决策树（Bayesian Decision Trees）与5层神经网络得到的基准实验结果，论文地址为：http://rdcu.be/cb58。该论文采用的是完整数据集，而非本数据集上传的前一百万条数据。