Searching for as yet undetected gamma-ray sources is a major target of the Fermi LAT Collaboration. We present an algorithm capable of identifying such type of sources by non-parametrically clustering the directions of arrival of the high-energy photons detected by the telescope onboard the Fermi spacecraft. n particular, the sources will be identified using a von Mises-Fisher kernel estimate of the photon count density on the unit sphere via an adjustment of the mean-shift algorithm to account for the directional nature of data. This choice entails a number of desirable benefits. It allows us to by-pass the difficulties inherent on the borders of any projection of the photon directions onto a 2-dimensional plane, while guaranteeing high flexibility. The smoothing parameter will be chosen adaptively, by combining scientific input with optimal selection guidelines, as known from the literature. Using statistical tools from hypothesis testing and classification, we furthermore present an automatic way to skim off sound candidate sources from the gamma-ray emitting diffuse background and to quantify their significance. The algorithm was calibrated on simulated data provided by the Fermi LAT Collaboration and will be illustrated on a real Fermi LAT case-study.

翻译：寻找尚未探测到的γ射线源是费米LAT合作组的主要目标之一。本文提出一种无需参数预设即可识别此类源的算法，该方法通过对费米卫星搭载望远镜探测到的高能光子到达方向进行非参数聚类实现。具体而言，通过调整均值漂移算法以适配数据的方向特性，利用冯·米塞斯-费舍尔核函数估计单位球面上的光子计数密度，从而实现对γ射线源的识别。该选择具有多项优势：既可规避将光子方向投影到二维平面时边界固有的处理难题，又能保证高灵活性。本文结合科学输入与文献中已知的最优选择准则，采用自适应方式确定平滑参数。此外，我们运用假设检验与分类等统计工具，提出一种自动方法以从γ射线弥散本底中筛选出可靠的候选源，并量化其显著性。该算法基于费米LAT合作组提供的模拟数据进行标定，并将在真实费米LAT案例研究中加以展示。