Projects
Here you can find my projects related to AI, Machine Learning, Deep Learning and Data Science.
- Fine-Tuned Object Detection (Jul 2025)
- Anime Faces Generator (May 2025)
- Translator Application (Jan 2025)
- Google Stock Price Forecasting (Apr 2025)
- Rice Image Classification (Feb 2025)
- Weather Application (Jan 2025)
- Password Generator Application (Jan 2025)
- Country Data (Dec 2024)
- Market Analyzing (Dec 2024)
- Customer Clustering (Nov 2024)
- Phone Price Prediction (June 2024)
- Personal Bank Loan (June 2024)
- Car Price Prediction (May 2024)
- Universities' World Ranking (Apr 2024)
- Microsoft Project (Jan 2023 - Sep 2023)
- Knowledge Management, Customer Relationship Management, Enterprise Resource Planning (Apr 2023 - May 2023)
- Management Principles and Organization Theory (Jan 2023 - Feb 2023)
- Road Transportation (Jan 2023)
- WBS (work breakdown structure) (Nov 2023 - Jan 2024)
- Hospital Ergonomics (Feb 2022 - Mar 2022)
- Improving Hospital Resource Allocation using Queuing Theory (Feb 2022 - Mar 2022)
- Introduce Lathe Machine (Mar2021 - Apr 2021)
- Demand in General Economy (Jan 2021)
I adapted the ultra-light YOLOv11-nano detector to a
multi-class traffic dataset. First, I organized images and annotations into training, validation
and test sets, then ran quick visual checks to ensure data quality. Leveraging
PyTorch and the Ultralytics API,
I fine-tuned the pretrained YOLOv11-nano model for 100 epochs at 640×640 resolution, closely monitoring loss,
precision, recall and mAP@0.5 from the training logs. After verifying strong quantitative metrics, I generated
sample images with overlaid bounding boxes to qualitatively validate detection performance. Finally, I ran
real-time detection on the dataset’s test video.
Repository on GitHub
· Notebook on
Kaggle
This project implement Generative Adversarial Networks
(GAN) to generate anime faces, one using TensorFlow and the other
PyTorch. It trains on the Anime Face Dataset from Kaggle, producing high-quality synthetic images after
40 epochs. The notebooks offer a practical, step-by-step guide to GAN implementation, making them great
for learning generative AI.
TensorFlow:
Repository on
GitHub · Notebook on Kaggle
PyTorch:
Repository on
GitHub · Notebook on Kaggle
The Translator app is a Vibe Coding project with minimalist web
application designed for real-time text translation. Built
with modern web technologies like TypeScript, Tailwind CSS and Vite, it offers a sleek and responsive
user interface.
You can clone and run the application locally through this
repository or access it directly via this link.
Repository on
GitHub
This project explores the application of Long Short-Term Memory
(LSTM) neural networks to forecast stock
price trends. By leveraging historical financial data, the model captures temporal patterns to predict
future price movements.
TensorFlow:
Repository on
GitHub · Notebook on Kaggle
PyTorch:
Repository on
GitHub · Notebook on Kaggle
This project classifies five rice varieties (Arborio, Basmati, Ipsala, Jasmine, Karacadag) using deep
learning. By training custom Convolutional Neural Networks (CNNs)
and fine-tuning ResNet50 on
75,000 images, it achieves 99% accuracy
through data augmentation, transfer learning and optimization (Adam, LR scheduling). Built with
TensorFlow
and PyTorch, the solution includes model deployment and performance visualization for scalable
agricultural
image analysis.
TensorFlow:
Repository on
GitHub · Notebook on Kaggle
PyTorch:
Repository on
GitHub · Notebook on Kaggle
A simple yet powerful weather application that retrieves and displays real-time weather information for
any
city worldwide using the OpenWeatherMap API. This project is designed to provide users with essential
weather details, including temperature, humidity, wind speed, and current weather conditions, in a clean
and
user-friendly interface.
You can clone and run the application locally through this
repository or access it directly via the internet using this link. The
application
is fully responsive and
works seamlessly across devices, ensuring a smooth experience for all users.
Repository on
GitHub
The Password Generator Web App is a simple yet powerful tool designed to create secure and customizable
passwords. Users can generate passwords in two modes: Simple and Complex. The Simple mode generates
passwords in the format ****-****-****, while the Complex mode
creates longer passwords in the format
****-****-****-****. Both modes include a mix of uppercase
letters, lowercase letters, numbers, and
special symbols (!@#$%&), ensuring strong and secure passwords. As it is an open source project you can
access it's code through the following repository:
Repository on
GitHub
This project applies Principal Component Analysis (PCA) to a
dataset containing country-level features.
The
goal is to reduce dimensionality while retaining most of the variance, enabling a more compact
representation of the data. PCA is used to:
Analyze the structure and variance of the dataset, visualize the data in reduced dimensions, evaluate
reconstruction accuracy after dimensionality reduction.
Repository on GitHub
This project is a comprehensive analysis of a financial market dataset, focusing on preprocessing,
visualization and predictive modeling. The objective is to explore various machine learning algorithms
to
predict stock prices and evaluate their effectiveness.
Repository on GitHub
· Notebook on
Kaggle
An Unsupervised Machine Learning Project which was about Clustering the customers using some
unsupervised
algorithms such as K-Means, DBSCAN and Gussian Mixture Model
(GMM), also on the GitHub
version
there is an emphasising on finding which one of the StandardScaler or MinMaxScaler is more appropriate
for
modelling. You can access the code on either GitHub or Kaggle.
Repository on
GitHub
· Notebook
on
Kaggle
The Phone Price Prediction project aims to classify mobile phones into different price ranges based on
their
technical specifications. Using machine learning algorithms, the project evaluates features like battery
power, RAM, processor cores and more to predict the price category of a phone. This project demonstrates
data preprocessing, feature selection and model optimization techniques to achieve accurate
predictions.
Repository on
GitHub
· Notebook on
Kaggle
In this ML project, I focus on predicting personal bank loan approval using Logistic
Regression, Naive Bayes and KNN algorithms. The goal is to analyze customer data such as age,
income,
and
education to predict whether a loan application will be approved or not.
Repository on
GitHub
· Notebook on
Kaggle
In this ML project, I used Linear Regression to predict car prices
based on various features such
as year, present price, kilometers driven, fuel type, seller type, and transmission. The model was
trained
to identify patterns in the dataset and accurately estimate car prices, helping to understand how
different
factors impact the value of a vehicle.
Repository on
GitHub
· Notebook on
Kaggle
In this project, I analyze universities across the world, utilizing various visualizations to highlight
patterns and distributions within the dataset.
Repository on
GitHub
· Notebook on
Kaggle
Participate in a group as a project manager to manage the project of constructing a building and setting up a knowledge base company.
Participate in a group to research and present, knowledge management, customer relationship management,
enterprise resource planning.
Presentation File
Participate in a group to present management, the importance of management and the levels of
management.
Presentation File
Participate in a group to present road transportation, different types of road transportation and also
mentioning some advantages and disadvantages in Transportation
subject.
Presentation File
WBS chart of the biggest project I've done so far in the Project Management subject at university. I participated in a group as a project manager to manage the project of constructing a building and setting up a knowledge base company.
A presentation for Human Factors Engineering subject which was
chosen because at that time the world
suffered from Corona virus.
Presentation File
Participate in a group to present how we can improve hospital resource allocation by using Queuing
Theory, which was chosen because at that time the world suffered from Corona virus.
Presentation File
Use an English video and translate it to Persian and present different part of lathe machine and how to machining in Manufacturing Methods.
Participating in a group to present the demand, demand curve and elasticity in General Economy.