Hi, I'm
Hamidreza Farhadipour
M.Sc. Data Science student & B.Sc. Statistics graduate — combining statistical rigor with machine learning, from Bayesian reliability to spatio-temporal graph neural networks.
- 18.74
- M.Sc. GPA — out of 20
- 6
- Academic Projects
- 7
- Certificates
- 2
- TA Courses
A little bit about who I am
Statistician by training, data scientist by craft, and a teacher at heart.
I am an M.Sc. Data Science student and B.Sc. Statistics graduate (Allameh Tabataba'i University), relying on statistical inference and predictive modeling to solve data-driven problems. I have hands-on research experience in advanced statistical analysis, machine learning, and deep learning — and I am comfortable running the full data-analysis cycle, from handling missing data to feature engineering, training, evaluation, and model interpretation, with an evidence-based and methodologically rigorous mindset.
My thesis focuses on enhancing reliability in traffic forecasting by integrating spatio-temporal graph neural networks with Bayesian approaches. Alongside research, I have served as a teaching assistant for Programming with R and Sampling Methods, and I enjoy collaborating on projects that extract valid insight from data, enable sharper decisions, and create measurable value.
Download full CVQuick facts
- Full nameHamidreza Farhadipour
- DegreeM.Sc. in Data Science (in progress)
- BirthplaceAbhar, Zanjan, Iran
- LanguagesPersian & Azerbaijani Turkish (native) · English (professional)
Research interests
Topics I actively explore, read about, and build on.
Graph Neural Networks & Traffic Forecasting
My thesis focus: spatio-temporal GNNs that model road-network structure for more reliable traffic prediction.
Bayesian Methods & Reliability
Probabilistic thinking for models you can trust — prior knowledge, uncertainty quantification, and reliability evaluation.
Time-Series Analysis
Forecasting with classical models (ARIMA/SARIMA) and neural approaches — with honest uncertainty and careful backtesting.
Machine Learning
From classical statistical learning to modern ensembles — building models that generalize, with rigorous validation.
Deep Learning & Computer Vision
Training modern CNN architectures for image classification and exploring transfer learning on real datasets.
Data Mining & Evolutionary Computation
Clustering, classification, and optimization with genetic algorithms and swarm intelligence — finding structure where nothing is labeled.
Things I have designed & built
Academic and course projects — from statistical modeling to deep learning.
Time-Series Modeling & Forecasting
Supervised by · Dr. Mohammadreza Salehirad
End-to-end forecasting study: decomposition and stationarity testing, ARIMA/SARIMA model selection, residual diagnostics, and accuracy benchmarking (RMSE/MAE) across candidate models.
Hybrid SQL + NoSQL Database Architecture
Supervised by · Dr. Mohammad Ebrahim Shiri
Designed and implemented a hybrid relational/document database architecture — mapping each workload to the right store and benchmarking cross-store queries for a realistic application scenario.
Comprehensive Data Analysis: Clustering & Classification
Supervised by · Dr. Vahid Rezaie Tabar
Full analytical cycle on a real dataset — preprocessing and missing-data handling, exploratory analysis, unsupervised segmentation, and supervised classification with model comparison.
Bayesian Email Spam Detection
Supervised by · Dr. Farzad Eskandari
Built a probabilistic spam filter from scratch — text preprocessing, Bag-of-Words / TF-IDF features, and Multinomial Naive Bayes evaluated with precision/recall trade-offs.
Pistachio Image Classification with Modern CNNs
Supervised by · Dr. Mohammadsharam Moein
Transfer-learning benchmark of modern CNN architectures for pistachio variety classification — augmentation strategies and a comparative evaluation of accuracy versus inference cost.
Evolutionary vs. Swarm Algorithms: A Comparative Study
Supervised by · Dr. Shokouh Shahbeik
Implemented Genetic Algorithms and Particle Swarm Optimization from scratch, comparing convergence, robustness, and solution quality on continuous and discrete optimization benchmarks.
Tools of the trade
Technologies and methods I use daily for research and development.
Programming & Databases
Machine Learning & Deep Learning
Statistics & Data Mining
Data Engineering & Visualization
Communication & Teaching
Education & experience
The path that shaped my academic and professional life.
Education
M.Sc. in Data Science
● Oct 2024 — PresentAllameh Tabataba'i University (ATU), Tehran
- GPA: 18.74 / 20
- Thesis: Enhancing reliability in traffic forecasting by integrating spatio-temporal graph neural networks with Bayesian approaches
B.Sc. in Statistics
Sep 2019 — Jul 2024Allameh Tabataba'i University (ATU), Tehran
- GPA: 15.98 / 20
- Core coursework: statistical inference, regression analysis, data mining, machine learning
Experience
Teaching Assistant — Programming with R
Graduate studiesAllameh Tabataba'i University
- Led hands-on R sessions and reviewed assignments for the Programming with R course.
- Supported students with data manipulation, visualization, and statistical computing in R.
Teaching Assistant — Sampling Methods
Graduate studiesAllameh Tabataba'i University
- Assisted the Sampling Methods course: prepared exercises on sampling designs and estimation, and graded coursework.
- Guided students through simple random, stratified, and cluster sampling concepts with practical examples.
Honors & certifications
Awards, scholarships, and professional certificates along the way.
Latest writing & notes
Short articles, tutorials, and notes on data science and ML.
A Gentle Introduction to Spatio-Temporal Graph Neural Networks
How GNNs encode road networks in space and time — and why they are changing traffic forecasting.
Bayesian Thinking for Trustworthy Machine Learning
Priors, posterior uncertainty, and what it means for a model to know what it doesn't know.
Time Series in R: From ARIMA to Modern Workflows
Practical forecasting recipes in R — the language I teach — with tidyverts and reproducible pipelines.
Let's work together
Have a research idea, a project, or just want to say hi? My inbox is always open.
💬 Always happy to talk about research collaboration, teaching, or data projects.