Tasks
1) Design of an improved DL/ML algorithm
2) Simulation of a monitoring centre
3) Optimization and improvement
4) Dissemination and exploitation
5) Quality and technical management
Task 1: Design of an improved DL/ML algorithm
1) Develop extensive set of ECG segments up to 40GB ECG data samples.
2) Feature engineering and training
3) GPU-based HPC environment, AutoKeras, TensorFlow
4) DL-based execution module expected to achieve over 92% accuracy.
Task 2: Simulation of a monitoring centre
Setup phase
1.a) 10K virtual patient that streams ECG
1.b) Developing docker instances and functions
Simulation phase
2.a) Serverless GCP or AWS
2.b) xAFCL orchestration engine of the workflow
2.c) HPC simulation
Evaluation phase
3.a) Measuring the response times, speedup and throughput
3.b) Comparison vs HPC cores
Task 3: Optimization and improvement
1) Implementing an improved DL model
2) Optimization and integration
3) Deployment to HPC environment
4) Measurement and evaluation of project KPIs
5) Preparing real-time heart monitoring center ready for market
Task 4: Dissemination and exploitation
1) Update of the dissemination and exploitation strategies,
2) Raising awareness about project goals,
3) Publishing project results, impact and benefits,
4) Publishing scientific results in highly ranked journals and conferences with golden access.
5) Update of the business model, strategy, and plans
Task 5: Quality and technical management
1) Planning and scheduling of project activities
2) Coordination and communication (weekly meetings)
3) Monitoring of project activities and results
4) Tendering procedure to select an HPC pro_vider
5) Quality management