AK

Amir Khoshnafs

Junior machine learning engineer focused on computer vision, deep learning, and reliable model evaluation.

About Me

Get to know the person behind the models - my background, interests, and what drives me in machine learning, computer vision, and statistics.

Amir Khoshnafs

Amir Khoshnafs

Machine Learning Engineer | Computer Vision | Deep Learning

My Story

I'm a junior machine learning engineer focused on computer vision and deep learning, with hands-on experience building modular Python projects in object detection, image segmentation, and video analytics.

My background in economics and ongoing graduate studies in statistics give me a strong quantitative foundation in regression, estimation, optimization, and reliable evaluation.

I'm especially interested in training behavior, model design, reproducible workflows, and turning machine learning ideas into practical, well-structured solutions.

Quick Facts

Izmir, Turkey
Graduate Studies in Statistics
B.Sc. in Economics
Machine Learning Engineer
Computer Vision & Deep Learning
Persian, Turkish, Azerbaijani, English

My Skills

A comprehensive overview of my technical background across programming, machine learning, computer vision, and workflow practices.

Programming & DBs

Python SQL Git Linux/Bash R MATLAB

Libraries & Frameworks

PyTorch YOLO OpenCV Pandas NumPy Scikit-learn Matplotlib Seaborn Scipy

Tools & Platforms

Reproducible Pipelines Model Evaluation Error Analysis Visualization Reporting

Featured Projects

Explore a selection of my computer vision and analytics projects, built around real-world data, practical problem solving, and clean end-to-end implementation.

Vehicle Speed Estimation

Vehicle Speed Estimation

A perspective-aware traffic analysis project that detects and tracks vehicles from video, then estimates their speed using computer vision techniques and scene calibration.

Python YOLO Tracking
U-Net Salt Segmentation

U-Net Salt Segmentation

A PyTorch-based image segmentation project that uses a U-Net architecture to identify salt deposits, with a focus on structured training, evaluation, and prediction quality.

Python PyTorch U-Net
Olist E-Commerce Analytics

Olist E-Commerce Analytics

This project analyzes real transactional data from the Brazilian e-commerce platform Olist using SQL and visualizes insights through a fully custom HTML dashboard.

Python SQL HTML

Want to see more?

Check out my GitHub for additional projects and contributions to the data science community.

View All Projects

My Journey

A timeline of my professional experience and educational background that shaped my foundation in machine learning, statistics, and quantitative analysis.

Experience

September 2024 - May 2025

Financial Educator / Analyst

Global Trading Insights · Remote

  • Produced and published over 30 educational videos on global trade, economics, and market dynamics.
  • Reviewed YouTube Studio metrics, including click-through rate, retention, and traffic sources, to improve topic selection and presentation.
  • Used performance insights to refine educational content strategy and strengthen audience engagement.
  • Reached international audiences through educational content on global trade, economics, and market dynamics.
July 2024 - August 2024

Finance Analyst Intern

Abbaszadeh Group · On-site

  • Supported expense and cash-flow tracking in Excel for monthly financial reporting.
  • Conducted market research on currency trends and trade operations, and organized reports for management review.
  • Assisted with organizing financial documents and supporting internal reporting workflows.

Education

2025 - Present

M.S. in Statistics

Ege University

Ongoing GPA: 3.88
Focus Areas
Statistical Learning Multivariate Statistics Computational Statistics Optimization
2020 - 2025

B.S. in Economics

Dokuz Eylul University

Graduated GPA: 2.66
Related Coursework
Econometrics Regression Analysis Statistics Financial Analysis

Get In Touch

I'm currently open to machine learning, computer vision, and data-focused opportunities. Whether you have a question, a collaboration idea, or a role in mind, my inbox is always open.

Email

amir_khoshnafs@yahoo.com

Phone

+90 (539) 544-5624

Location

Izmir, Turkey