High energy physics (HEP) experiments are collecting unprecedented amounts and variety of data. Major upgrades of the current HEP experimental setups already in progress will further increase the volume and complexity of this data.
This new territory of extremely large data is inspiring and challenging physicists to devise an ever growing number of new and more advanced analysis techniques, which increasingly include elaborate procedures such as machine learning methods. All these developments, which will have a growing impact on HEP in the coming years require fast, efficient and flexible analysis techniques and incorporation of state-of-the-art hardware infrastructures. Following and adapting recent advances in data science, computing, and related fields, and closely interacting with field experts has already led to a significant leap in HEP, in particular, in data analysis. It is therefore worthwhile to concentrate further on inventing new data handling and analysis systems methodologies based on formal developments in these fields, with the aim to facilitate accommodating effective and sophisticated techniques into HEP data analysis.
The focus of this article collection is on contributions from the HEP community which currently explore this area of research. We welcome contributions from a wide range of technical subjects, which can be classified into the following main areas:
- Innovative designs towards user-friendly, preservable analysis frameworks and dedicated languages for description of analysis algorithms.
- Developments towards fast analysis on high performance computing centers (HPCs) using modern parallelization techniques and benchmarking the performance of HPCs with respect to traditional Grid infrastructures.
- Developments of analysis infrastructures that can be efficiently interfaced to machine learning applications, optimization pipelines, and that can benchmark the performance of different data formats in terms of long-time persistence, I/O performances and flexibility.
High energy physics (HEP) experiments are collecting unprecedented amounts and variety of data. Major upgrades of the current HEP experimental setups already in progress will further increase the volume and complexity of this data.
This new territory of extremely large data is inspiring and challenging physicists to devise an ever growing number of new and more advanced analysis techniques, which increasingly include elaborate procedures such as machine learning methods. All these developments, which will have a growing impact on HEP in the coming years require fast, efficient and flexible analysis techniques and incorporation of state-of-the-art hardware infrastructures. Following and adapting recent advances in data science, computing, and related fields, and closely interacting with field experts has already led to a significant leap in HEP, in particular, in data analysis. It is therefore worthwhile to concentrate further on inventing new data handling and analysis systems methodologies based on formal developments in these fields, with the aim to facilitate accommodating effective and sophisticated techniques into HEP data analysis.
The focus of this article collection is on contributions from the HEP community which currently explore this area of research. We welcome contributions from a wide range of technical subjects, which can be classified into the following main areas:
- Innovative designs towards user-friendly, preservable analysis frameworks and dedicated languages for description of analysis algorithms.
- Developments towards fast analysis on high performance computing centers (HPCs) using modern parallelization techniques and benchmarking the performance of HPCs with respect to traditional Grid infrastructures.
- Developments of analysis infrastructures that can be efficiently interfaced to machine learning applications, optimization pipelines, and that can benchmark the performance of different data formats in terms of long-time persistence, I/O performances and flexibility.
