Parisa Shahabinejad

Wireless Sensor Network Optimization Using Evolutionary Algorithms

Implemented an evolutionary algorithm to optimize the cost and service quality of a city-wide communication network by adjusting the number, location, and transmit power of sensor nodes.

Python Evolutionary Algorithms Optimization Wireless Sensor Networks

Game-Playing Intelligent Agent with Reinforcement Learning

Implemented reinforcement learning algorithms, including Value-Iteration and Q-Learning, for a 2D grid world Markov Decision Process resembling a Pac-Man game. Also applied the Mini-Max algorithm and common path-planning techniques such as A*, Dijkstra, and bidirectional search.

Python Reinforcement Learning Mini-Max MDP Q-Learning Algorithms

Comparison of MLP and RBF as Classifiers on CIFAR-10

Implemented Multilayer Perceptron (MLP) and Radial Basis Function (RBF) neural networks from scratch in Python and used ResNet-34 as a feature extractor. Evaluated and compared the classification accuracy of the two networks on the CIFAR-10 dataset.

Python Neural Networks CIFAR-10 Classification MLP RBF ResNet-34

Neural Architecture Search for CIFAR-10 Image Classification Using Evolutionary Algorithms

Applied evolutionary algorithms to automate deep neural network design for image classification. The network architecture was optimized by selecting the best feature extractor, number of hidden layers, number of neurons, and activation function from a predefined search space.

Python Neural Networks NAS Evolutionary Algorithms CIFAR-10

Clustering of Image Feature Vectors Using Self-Organizing Map

Applied a self-organizing map network to cluster feature vectors extracted from ResNet-34 on the CIFAR-10 dataset. Explored different network configurations, such as one-dimensional and two-dimensional topologies and neighborhood radius, to analyze the clustering results.

Python Self-Organizing Map Clustering CIFAR-10

Travel Insurance Prediction Model

Developed a predictive model to forecast travel insurance subscriptions using MLP Classifier and Logistic Regression after data analysis and preprocessing in Python.

Python EDA MLP Classifier Logistic Regression Machine Learning

Behavioral Data Clustering and Gender Correlation Analysis

Clustered behavioral data into two groups, regardless of gender, and evaluated cluster consistency with gender division using silhouette and Davies-Bouldin scores.

Python K-Means EDA Clustering Silhouette Score Davies-Bouldin Score

Applicant Project Score Prediction Based on Technical Test Results

Developed a model to predict project scores based on technical aptitude test results. Cleaned and normalized the data, then applied regression analysis to forecast grades.

Python EDA Linear Regression MLP Regressor Machine Learning

Persian News Article Text Classification Using Pre-trained BERT Model

Developed a text classification model to categorize Persian news articles into seven topics by fine-tuning the Hugging Face pre-trained models.

NLP Text Classification BERT

Jobonja - A Dynamic Job Portal Website

Developed a Next.js-powered website that connects job seekers and employers, implementing design patterns, unit tests, CI/CD, and frontend optimization techniques for enhanced performance as part of the Object-Oriented Design course project.

Next.js Jest Performance Optimization Design Patterns CI/CD Bootstrap

Heifes - E-commerce Platform Development

Developed Heifes, a Next.js-based e-commerce platform, connecting users with stores having surplus unsold food. Implemented with Scrum, detailed UML diagrams, 60% test coverage via Jest, and Cypress integration tests.

Next.js React MongoDB TypeScript Tailwind Jest Cypress Ant Design Scrum

Heifes Website