Raziyeh Dehbozorgi

Ph.D. in Applied Mathematics — PDE&IE based-Modelling, Numerical Analysis & Deep Generative Models

Explore my work

Education

Ph.D. Applied Mathematics (IUST), M.Sc. (Rafsanjan), B.Sc. (Shiraz)

Contact

r.dehbozorgi2012@gmail.com · +98 910 007 4178

Research Interests

PDEs, numerical analysis and deep generative models

Publications

7 Q1 journal papers & 7 conference papers

Experience

6+ years postdoctoral research at IASBS

Toolkit

Languages, frameworks & scientific computing tools

  • Numerical Analysis & PDEs

    — MATLAB, FreeFem++, Mathematica
  • Scientific Computing

    — Python, Julia, C/C++, R
  • Deep Generative Models

    — PyTorch, TensorFlow, Diffusion Models

About Me

Applied mathematician and deep learning researcher, six-plus years at the intersection of scientific computing and generative AI.

Raziyeh Dehbozorgi

Core expertise spans PDE&IE-based modelling, numerical analysis, and deep generative models — including CGANs and diffusion models such as DDPM, DDIM, LDM, DBSM and DBM — applied to inverse problems, computational pathology, and healthcare imaging.

Author of 7 peer-reviewed journal papers (5 of them Q1-ranked) and 7 conference papers, with 2 further works under review. Now seeking to bring this scientific ML/DL background to applied research, R&D, or scientific computing roles in industry.

Ph.D., Applied Mathematics

Iran University of Science and Technology · 2014–2018

M.Sc., Applied Mathematics

Rafsanjan University · 2010–2013

B.Sc., Applied Mathematics

Shiraz University · 2005–2009

Academic & Teaching Experience

Postdoctoral research, lecturing and teaching assistantships across six-plus years.

2024 – 2025

Postdoctoral Researcher, Institute for Advanced Studies in Basic Sciences (IASBS)

Designed a Deep Ritz learning framework for solving elliptic PDEs using finite-element datasets (FreeFem++) to constrain and validate neural network solutions, tested on Poisson and Schrödinger equations against classical FEM baselines.

2019 – 2022

Postdoctoral Researcher, IASBS

Developed an adaptive hp-collocation method for nonlinear Abel inverse problems (existence, uniqueness, stability and error analysis in weighted Sobolev space), and engineered a Galerkin FEM/SEM solver for nonlinear fractional boundary value problems.

2022 – 2024

Part-time Lecturer, Azad Universities, Tehran

Taught 6 core mathematics courses (Calculus I/II, ODEs, Linear Algebra, Numerical Analysis, Probability, Applied Mathematics in Finance) to undergraduate cohorts of 40–60 students.

2018 – 2019

Part-time Lecturer, University of Applied Sciences & Technology

Taught Calculus I/II, ODEs and Numerical Analysis to undergraduate cohorts of 40–60 students.

2014 – 2015

Teaching Assistant, Iran University of Science and Technology

Supported instruction across 4 advanced courses and co-supervised at least 7 master's students through completion of their thesis research.

2018 – present

Freelance Programming & Tutoring

Project-based instruction, research and consulting in Python, MATLAB, C++, Java, ML/DL, data analysis, Power BI and Arduino.

Professional Activities & Recognition

Reviewing, membership, awards and selected AI projects.

Reviewing & Membership

Journal reviewer for AMS MathSciNet, Journal of Computational and Applied Mathematics, JAMP, FILOMAT, Mathematical Modeling Journal and Mathematical Science Journal. Conference reviewer, 51st Annual Iranian Mathematics Conference (2021). Member, American Mathematical Society & Iranian Mathematical Society. Book editor, "Introduction to Mathematics" (Prof. Sepehri, Shiraz University).

Awards & Achievements

First-ranked graduate student across all degrees (GPA ≈ 4.0). Rank 23 of ~8,000 in the national Ph.D. entrance exam; member, Distinguished Talent Society (2018). Ph.D., M.Sc. and B.Sc. government scholarships (2005–2018).

Selected AI Projects

Segmentation of brain regions for Multiple Sclerosis diagnosis via VGG19, VGG16 and ResNet50. Signal processing for speech using FFT and wavelet transforms. Image retrieval using invariant Krawtchouk moments.

Certifications, Courses & Seminars

Continued learning across machine learning, generative models and applied AI.

8
Q1 Journal Papers
8
Conference Papers
7
Years Postdoc Research
3
Papers Under Review

Research Interests

The connective tissue between rigorous numerical analysis and modern generative modelling.

Numerical Analysis / Approximation Theory Deep Generative Models Finite / Spectral Element Methods Stochastic Differential Equations PDE-Based Modelling Nonlinear, Nonlocal & Neural Operators Fractional PDEs / Integral Equations Inverse Problems Optimal Control / Optimization Computer Vision
∇²u = f ∑∈𝚫

Toolkit

Languages, frameworks and scientific-computing tools used across research and teaching.

Languages & ML Frameworks

Python PyTorch TensorFlow Scikit-Learn Pandas NumPy PySpark MATLAB Java C / C++ Julia R html css

Scientific & Other Tools

Mathematica FreeFem++ SQL Git Power BI Arduino Assembly

Publications

7 Q1 journal papers, 7 conference papers, and ongoing work in generative modelling for science.

Conference
Hybrid functions method for solving systems of integral equations

49th Annual Iranian Mathematics Conference, Tehran, 2018

Conference
Numerical solution of weakly singular integro-differential equations in anomalous diffusion

49th Annual Iranian Mathematics Conference, Tehran, 2018

Conference
Numerical solution of Riccati and nonlinear Volterra integral equations using operational vectors

49th Annual Iranian Mathematics Conference, Tehran, 2018

Conference
Optimal control constrained by parabolic evolution equations via wavelet functions

9th Seminar on Nonlinear Analysis, Qazvin, 2018

Conference
hp-collocation method for optimal control problems in vulnerable computer systems

9th Seminar on Nonlinear Analysis, Qazvin, 2018

Under Review
Comparative analysis of diffusion generative models in computational pathology

Journal of Pathology Informatics — major revision, 2025

Iran University of Science and Technology IASBS Rafsanjan University Shiraz University Azad University, Tehran

Get in Touch

Open to applied research, R&D and scientific computing roles. Reach out directly, or use the form below.

Your message has been sent. Thank you!