Ramtin Ardeshirifar's repositories
Attention-Based-LSTM-Network-for-Predicting-Times-Series
Forex price movement forecast
Predicting-Stock-Prices-Using-FB-Prophet
Prophet is a procedure for forecasting time series data based on an additive model where non-linear trends are fit with yearly, weekly, and daily seasonality, plus holiday effects. It works best with time series that have strong seasonal effects and several seasons of historical data. Prophet is robust to missing data and shifts in the trend, and typically handles outliers well. In this notebook I'm going to try forecasting Google stock price using facebook's prophet model.
Traffic-forecasting-using-different-methods
Time series prediction using deep learning
Predicting-YouTube-Dislikes-using-Machine-Learning
I used Catboost for training a model on the numerical features of every YouTube video (e.g., the number of views, comments, likes, etc.) along with sentiment analysis of the video descriptions and comments using the VADER sentiment analysis model.
Credit-Card-Fraud-Detection
In this project, I explore different methods for detecting credit card fraud transactions; including using the Catboost algorithm with undersampling & oversampling methods, and using an almost new approach, by using deep learning and autoencoder.
Name-Nationality-Classification-with-Machine-Learning
Predicting nationality only by using people's names with machine learning.
Word-Clouds-With-Sentiment-Analysis-of-the-Most-Recent-Tweets
Downloaded tweets from the most popular news agencies and extract keywords from them. In the next steps, I plotted a word cloud and did a sentiment analysis for tweets that have the keywords.
Detecting-Machine-Generated-Text
The findings of this research reveal several intriguing disparities between human and AI text generation. I demonstrated that these differences could be successfully utilized by classifiers to distinguish between human and AI-generated text.
Automated-Design-of-Symmetric-Autoencoders-Using-Genetic-Algorithms
This project presents a novel approach for optimizing the architecture of symmetric, undercomplete autoencoders for dimensionality reduction using a genetic algorithm.