The following were selected by DAIS ITA leaders spanning the five-year program from key achievements nominated by DAIS Technical Area Leaders across eight different categories core to the DAIS values and mission.

Cross National Collaboration

  • Adaptive Federated Learning in Resource Constrained Edge Computing Systems (IBM US, Imperial, PSU, ARL)
  • Interpretability of Deep Learning Models: A Survey of Results (IBM US, IBM UK, Cardiff, UCLA, UCL, ARL, IBM UK, Ozyegin University, BAE)
  • Network Research Collaboration (Imperial, Yale)

Interdisciplinary Collaboration

Straddling the goals of collaboration & advanced science, and speaks to program breadth

  • Applying AI to criminal and terrorist networks (Penn State, IBM US, IBM UK)
  • Network Motif Analysis (Cardiff, Penn State, UMass, IBM UK)
  • Semantic Vectors (ARL, IBM UK, Cardiff)

Scientific discovery / quality

  • Adaptive Federated Learning in Resource Constrained Edge Computing Systems (IBM US, Imperial, PSU, ARL)
  • Evidential Deep Learning to Quantify Classification Uncertainty (ARL)
  • A Scalable Vector Symbolic Architecture Approach for Decentralized Workflows (Cardiff, IBM UK, IBM US)
  • The evolution of strongly-held group identities through agent-based cooperation (Cardiff, Yale)

Blue sky exploration

Also known as ‘scientific risk’

  • Towards spike-based machine intelligence with neuromorphic computing (Purdue)
  • Coalition Situational Understanding Facilitated By Vector Symbolic Architecture (Cardiff, IBM UK. UCLA, BAE)
  • DeepProbCEP: A Hybrid Neuro-Symbolic Approach for Complex Event Processing in Noisy and Adversarial Settings (Cardiff, UCLA)

Military impact

  • Embedding Dynamic Knowledge Graphs based on Observational Ontologies in Semantic Vector Spaces (UK transition)
  • Machine learning algorithms promise better situational awareness ARL news
  • Research enhances Soldier decision-making ARL news

Commercial prosperity

  • Federated Learning in IBM Cloud Pak 3.5
  • FastLAS - Mark Law from Imperial has set up his own ILASP business, contracting to Dstl to advance ILASP and FastLAS for military operations
  • Robust CoreSet Construction in IBM Research Early Access on Edge AI SDK

Social Good / Community Impact

  • PhD Student Cohort (50+ students across lifetime of program)
  • Summer Intern Placement Program (25+ students spent summer in rotation at IBM)
  • USMA Cadets (Hosted visits from 2016-2019 for 2-4 visiting cadets at Hursley per year)

Student-led Achievement (training next-generation researchers)

  • Harrison Taylor/Alun Preece Example Paper - VADR: Discriminative Multimodal Explanations for Situational Understanding
  • Tiffany Tuor/Kin Leung Example Paper - Distributed Machine Learning in Coalition Environments: Overview of Techniques
  • James Ashford/Roger Whitaker Example Paper - Understanding the Signature of Controversial Wikipedia Articles through Motifs in Editor Revision Networks (www conf)