Guideline

Introduction

Welcome to my portfolio! My name is Masih Moafi, and I'm passionate about leveraging AI and machine learning applications to develop innovative, cutting-edge solutions. My primary focus areas are financial markets, medicine, and unsupervised learning. Through this portfolio, I aim to showcase my work and projects across various domains

Projects

1. Financial Market Analysis

1.1 Time-Series Analysis for Bitcoin

• Description: This project involves the analysis of Bitcoin's(or any other chart) time-series data using LSTM and ResNet in conjunction. Later, 2 trading strategies are implemented. One of which is profitable.
• Algorithms: ResNet-LSTM

1.2 Clustering 100 Cryptocurrencies

• Description: clustering 100 crypto currencies using K-means algorithm.
• Algorithms: K-Means Clustering

2. Medical Image Segmentation

• Description: In this project, I have used the U-Net architecture to perform image segmentation in medical diagnosing. It focuses on multiple classification tasks related to medical images.
• Algorithms: U-Net

3. Speech Recognition with CNN-Transformers

• Description: This project explores speech recognition using a combination of CNN and Transformers. It aims to achieve accurate speech-to-text conversion.
• Algorithms: CNN-Transformers

4. NLP: Sentiment Analysis of Tweets

• Description: This project involves sentiment analysis of tweets using logistic regression and ensemble tree models. It was a Kaggle competition that required analyzing and classifying the sentiment of tweets.
• Technologies: Logistic Regression, Ensemble Trees, NLP

5. Transfer Learning Projects

5.1 Text Generation using GPT-2

• Description: This project utilizes GPT-2 for text generation tasks. It showcases the generation of customized and coherent text based on pre-trained models.
• Technologies: LLMs(GPT-2)

5.2 Style Transfer using Transfer Learning

• Description: This project demonstrates style transfer by combining the style of one image with the content of another. It showcases the power of transfer learning techniques.
• Technologies: Transfer Learning, Style Transfer, Tensorflow

6. Movie Recommendation System

• Description: In this project, I have implemented a movie recommendation system using content-based clustering algorithms. It helps users discover relevant movies based on their preferences.
• Technologies: K-Means Clustering, Content-based Filtering

7. Predicting Housing Prices

• Description: This project tackles a Kaggle competition that involves predicting housing prices. It focuses on handling messy, missing, and skewed data and employs XGBoost for accurate predictions.
• Technologies: XGBoost, Regression, data preprocessing

8. Work Hour Analysis

• Description: This project involves analyzing work hours using linear regression and polynomial regression models. It showcases the ability to analyze and predict work patterns.
• Algorithms: Linear Regression, Polynomial Regression

Contact

To get in touch with me or discuss potential collaborations, please feel free to email me at [email protected].