More Like this

1. Direct photon production and PDF fits reloaded

   https://doi.org/10.1140/epjc/s10052-018-5944-4  

   Campbell, John M.; Rojo, Juan; Slade, Emma; Williams, Ciaran (June 2018, The European Physical Journal C)
2. Photon PDF within the CT18 global analysis

   https://doi.org/10.1103/PhysRevD.105.054006  

   Xie, Keping; Hobbs, T. J.; Hou, Tie-Jiun; Schmidt, Carl; Yan, Mengshi; Yuan, C.-P. (March 2022, Physical Review D)
3. A Learning Theoretic Perspective on Local Explainability (pdf)

   Jeffrey Li, Vaishnavh Nagarajan (March 2021, Proceedings of the Ninth International Conference on Learning Representations)

   In this paper, we explore connections between interpretable machine learning and learning theory through the lens of local approximation explanations. First, we tackle the traditional problem of performance generalization and bound the test-time accuracy of a model using a notion of how locally explainable it is. Second, we explore the novel problem of explanation generalization which is an important concern for a growing class of finite sample-based local approximation explanations. Finally, we validate our theoretical results empirically and show that they reflect what can be seen in practice. 

   more » « less
4. Understanding PDF uncertainty in W boson mass measurements*

   https://doi.org/10.1088/1674-1137/ac930b  

   Gao, Jun; Liu, Dianyu; Xie, Keping (December 2022, Chinese Physics C)

   Abstract We study the dependence of the transverse mass distribution of charged leptons and the missing energy on parton distributions (PDFs) adapted toWboson mass measurements at the CDF and ATLAS experiments. We compare the shape variations of the distribution induced by different PDFs and find that the spread of predictions from different PDF sets can be significantly larger than the PDF uncertainty predicted by a specific PDF set. We suggest analyzing the experimental data using up-to-date PDFs to gain a better understanding of the PDF uncertainties inWboson mass measurements. We also perform a series of Lagrange multiplier scans to identify the constraints on the transverse mass distribution imposed by individual data sets in the CT18 global analysis. In the case of the CDF measurement, the distribution is mostly sensitive tod-quark PDFs in the intermediatexregion, which are largely constrained by DIS and Drell-Yan data on deuteron targets and Tevatron lepton charge asymmetry data. 

   more » « less
5. ChemScraper: leveraging PDF graphics instructions for molecular diagram parsing

   https://doi.org/10.1007/s10032-024-00486-7  

   Shah, Ayush Kumar; Amador, Bryan; Dey, Abhisek; Creekmore, Ming; Ocampo, Blake; Denmark, Scott; Zanibbi, Richard (July 2024, International Journal on Document Analysis and Recognition (IJDAR))

   Most molecular diagram parsers recover chemical structure from raster images (e.g., PNGs). However, many PDFs include commands giving explicit locations and shapes for characters, lines, and polygons. We present a new parser that uses these born-digital PDF primitives as input. The parsing model is fast and accurate, and does not require GPUs, Optical Character Recognition (OCR), or vectorization. We use the parser to annotate raster images and then train a new multi-task neural network for recognizing molecules in raster images.We evaluate our parsers using SMILES and standard benchmarks, along with a novel evaluation protocol comparing molecular graphs directly that supports automatic error compilation and reveals errors missed by SMILES-based evaluation. On the synthetic USPTObenchmark, our born-digital parser obtains a recognition rate of 98.4% (1% higher than previous models) and our relatively simple neural parser for raster images obtains a rate of 85% using less training data than existing neural approaches (thousands vs. millions of molecules). 

   more » « less