Data for: Detecting Explosive Trace Dataset using CNN

The dataset was originally numerical dataset for Detecting Explosives by PGNAA using KNN. Due to the need to train the data for trace element detection using CNN, it was important to modify the data by converting them to 2D images. CNN performs better with images than numerical values.
Out of 65, 000 data samples, 1000 points were used. Each data point had only two features, explosive and non-explosive.

该数据集最初为基于PGNAA的爆炸物检测数值数据集，旨在通过KNN算法实现。鉴于训练数据以用于CNN进行微量元素检测的必要性，对数据进行修改，将其转换为二维图像变得尤为关键。相较于数值数据，卷积神经网络在图像处理方面表现更为卓越。在65,000个数据样本中，选用了1,000个数据点。每个数据点仅包含两个特征：爆炸物和非爆炸物。