Predictive Model Development
Budget: £20 – £250 GBP
I am looking for an experienced researcher or engineer to develop a data-driven dynamic model using system identification methods. The goal is to design a model that can:
• Learn from available input–output data
• Be validated against existing datasets
• Predict system outputs accurately for new, unseen inputs
Preferred approaches include:
• Sparse Identification of Nonlinear Dynamics (SINDy)
• or Subspace Model Identification
Expected Deliverables:
1. A working predictive model (MATLAB)
2. Documentation of methodology and implementation steps
3. Demonstration of model performance on both training and new input data
⸻
Required Skills
• Strong background in system identification, control systems, or machine learning
• Practical experience with SINDy, Subspace Identification, or related techniques
• Proficiency in MATLAB
• Ability to explain methods clearly and provide reproducible code
• Preferably familiar with electric motors, but this is not essential
• Learn from available input–output data
• Be validated against existing datasets
• Predict system outputs accurately for new, unseen inputs
Preferred approaches include:
• Sparse Identification of Nonlinear Dynamics (SINDy)
• or Subspace Model Identification
Expected Deliverables:
1. A working predictive model (MATLAB)
2. Documentation of methodology and implementation steps
3. Demonstration of model performance on both training and new input data
⸻
Required Skills
• Strong background in system identification, control systems, or machine learning
• Practical experience with SINDy, Subspace Identification, or related techniques
• Proficiency in MATLAB
• Ability to explain methods clearly and provide reproducible code
• Preferably familiar with electric motors, but this is not essential