Ph.D. position at TECNALIA for the IN-DEEP European Doctoral Network Real-time inversion using self-explainable Deep Learning driven by expert knowledge

Reference number: DC3

PhD research topic: Uncertainty-aware deep learning for parametric PDEs modelling tumor growth.

Host institution: Tecnalia Research & Innovation, Spain

PhD Enrolment: University of the Basque Country (UPV/EHU), Bilbao, Spain

Main Supervisor: Prof. Dr. Javier del Ser, Fundación Tecnalia Research & Innovation (Tecnalia), Derio, Spain, javier.delser@tecnalia.com

Co-supervisor: Prof. Dr. Alessandro Reali Universita Degli Studi Di Pavia (UniPV), Pavia, Italy, alessandro.reali@unipv.it

Scientific tasks

Overall research topic: Uncertainty, Explainability and Active Learning methodologies for Physics-/Domain-expertise Driven Neural Networks

• To gauge the impact of different methodologies to incorporate physics-/domain-expertise on the epistemic uncertainty of neural networks.


• To align the knowledge modeled by a neural network to explanations verified or produced by experts (e.g. a manual annotation of the important regions of an input image).


• To reduce the uncertainty propagated to the output of a network by devising automatic data augmentation methodologies that are driven by the analysis of the quantified uncertainty in goal 1.


• To showcase the above goals in synthetic and real-world modeling tasks: road traffic characterization and/or tumor growth.

For more information on the IN-DEEP Doctoral Network, please visit https://www.in-deep.science

Expected outcomes

1. Physics-aware neural network algorithm with quantifiable uncertainty propagation.


2. Active learning strategies that leverage expert knowledge to reduce the uncertainty in the proposed neural network model.


3. Methods to align the knowledge of the neural network with external expert annotation.


4. Application of the proposed method to parametric models of tumor growth and/or road traffic models.


5. 2+ peer-reviewed publications.


6. 2+ participation in relevant international conferences.

Eligibility Criteria

• Mobility: At the time of recruitment, the researcher must not have resided or carried out his/her main activity (work, studies, etc.) in Spain for more than 12 months in the 36 months immediately before the recruitment date. Time spent as part of a procedure for obtaining refugee status under the Geneva Convention or compulsory national service are not considered.


• The candidate must be, on the date of recruitment, a doctoral candidate (e. not already in possession of doctoral degree). Researchers who have successfully defended their doctoral thesis but who have not yet formally been awarded the doctoral degree will not be considered eligible.


• The candidate must agree to work exclusively for the action.

Contract: Full-time contract

Duration: 36 months, including 2 secondments at other consortium members’ premises

Estimated starting date: September 1st, 2024 (at the latest)

Salary: The gross salary amounts to €28.000 (estimated) per year. The candidate will also receive a mobility allowance of €600 euro per month, and, where applicable, a family allowance of €660 per month and a special needs allowance.

Evaluation criteria

• Step 1:
  
  • Academic performance during undergraduate studies - 30 points.
  
  
  • Research experience in the topic of the call, including publications, projects, and internships - 30 points.
  
  
  • Awards, honors, other significant roles, and achievements as a student - 10 points.
  
  
  • Additional coursework, certifications, training programs, continuous learning - 5 points.
  
  


• Step 2: Only for those scoring 50 or above in step 1:
  
  • Interview to assess communication skills, initiative, critical thinking, and motivation to pursue a PhD. - 15 points.
  
  
  • Letters of recommendation - 10 points.
  
  

In case of candidates achieving equal scores, priority will be given to female applicants.

Application: Please send an email with your CV, a letter of interest (up to one page), and the contact information of two referees who can provide letters of recommendation to javier.delser@tecnalia.com

Women are encouraged to apply!

This Project has received funding from the EU’s Horizon Europe research and innovation programme under the Marie Sklodowska-Curie GA No 101119556

Specific requirements

• Bachelor’s degree in computer science, Engineering, Physics, Mathematics, or any other related fields.


• Studies that grant access to a Ph.D. Program by September 1st, 2024.


• Experience in Machine Learning and Data Science.


• Previous experience programming in scientific computing.


• Language: Excellent command of English, together with good academic writing and presentation skills. Knowledge of Spanish is advisable to foster interaction with other lab mates, but not mandatory.

Preferred skills

• MSc in Artificial Intelligence, Data Science, Machine Learning, or related topics will be very positively welcome.


• Familiar with neural networks and deep learning algorithms.


• Experience with numerical methods for partial differential equations.


• Notions on uncertainty estimation (e.g. Bayesian ML, evidential DL, Monte Carlo dropout) and Partial Differential Equations (PDE, ODE, …) are warmly welcome.


• Programming skills: Python, Tensorflow, Pytorch, Scikit-learn, Latex. Linux at user level.


• Social skills: open-minded, resilient to criticism, objective-driven, social, team-work-oriented.


• Able to work independently and flexibly, taking initiative where required.