Find more staff
Role:
Department staff:
Research staff:
- Renewable Energy
- Applied Machine Learning
- Statistical Physics
- Nonequilibrium Dynamics
- Mathematical &
- Computational Modelling in Finance and Engineering
- Stochastic Modelling
- Complex Systems
- Climate Data Analysis
- Data Science
- Artificial Intelligence (AI)
- Complex Networks
Teaching staff:
- Qualifications:PhD in Statistical Physics from the Universitat Politècnica de Catalunya (Barcelona, Spain) BSc in Statistical Physics from Instituto Balseiro (Bariloche, Argentina)
- Position:Senior Lecturer in Data Science
- Department:FET - Computer Science and Creative Technologies
- Telephone:+441173283289
- Email:Ignacio.Deza@uwe.ac.uk
-
Social media:
About me
I am a Senior Lecturer in Data Science and Deputy Programme Leader of the BSc in Data Science at the University of the West of England (UWE), Bristol. My research bridges Statistical Physics, Data Science, and AI, focusing on developing structural alternatives to standard machine learning when dealing with complex, noisy, and highly skewed datasets.
Following a PhD in Statistical Physics from the Universitat Politècnica de Catalunya as a European Horizon 2020 fellow, I have worked across European research environments to translate fundamental science into applied data solutions. My published work spans climate dynamics, renewable energy, subscription markets, and aerospace. These domains are united by a common challenge: the need for bespoke statistical architectures where conventional, off-the-shelf AI fails.
A key output of this research is my open-source development, including qNoise, a software package for generating non-Gaussian coloured noise. This is used internationally to support research in stochastic modelling, signal analysis, and algorithm stress-testing.
At UWE, I lead the Fundamentals of Machine Learning module and drive the BSc Data Science curriculum to ensure rigorous theoretical foundations meet practical industry relevance.
Industrial Impact & Technical Advisory
I am highly active in knowledge exchange, bringing together expertise from physics and advanced statistics to tackle genuinely interdisciplinary problems. I provide technical advisory and stochastic auditing for industrial predictive systems. Organisations or researchers facing challenges with sparse measurements, heavy-tailed risk, or the failure of standard predictive models can review my applied methodological frameworks and open-source case studies via my GitHub Portfolio.
Area of expertise
-
Rare Event & Extreme Risk Forecasting: Predicting infrequent but critical events (equipment failure, rapid subscriber churn, environmental anomalies) where standard algorithms systematically under-evaluate risk.
-
Model Failure Diagnostics & Predictive Reliability: Identifying why machine learning pipelines achieve high accuracy in training but fail in production, specifically addressing severe class imbalance and time-series forecasting errors.
-
Small Data & Sparse Modelling: Building robust predictive systems when data is limited, noisy, or highly skewed, bypassing the massive data requirements of standard deep learning tools.
-
Root Cause Analysis in Complex Systems: Moving beyond simple correlation to extract true causal drivers and structural dependencies within messy, real-world industrial and IoT data networks.
-
Heavy-Tailed Risk & Outlier Management: Correcting the predictive blind spots of standard Gaussian assumptions in systems that exhibit extreme outliers and power-law dynamics.
-
Curriculum Design & Academic Leadership: Deputy Programme Leader for BSc Data Science, embedding these rigorous, problem-led methodologies into the foundations of applied statistics and AI education.
Publications
Publications loading...