Sina Emami

By passing the channel information such as username and id, this script would scrap all the posts in the channel. First, you need to pass your account information so it can connect to your telegram account.

• Link: GitHub
• Date: Aug 2021
• Technology: Python, Telethon

My personal web page which is one of my very few front projects. I download a template and customize it with adding features to it such as animation effect, changing the structure and color of in the way i liked.

• Link: GitHub
• Date: Sep 2021
• Technology: CSS, HTML, javascript, bootstrap

The goal for ISIC 2019 is classify dermoscopic images among nine different diagnostic categories I have developed a model to classify the images using CNN architecture and softmax activation function. Also, I have used regularization techniques to reduce the overfitting issue.

• Link: GitHub
• Date: Dec 2020
• Technology: Python, TensorFlow, Keras, Pandas, Numpy

Developed a bot to automatically copy the trade of a trader for all its followers in their accounts with APIin Binance exchangeless than a second. Also add some features to the trading so that some parts of the buying and selling for the main master would be done automatically by the bot.
Wrote unit-test for it with 89% code coverage.

• Client: Nurafarin
• Date: April 2021
• Technology: Python, Flask, SQLite, Redis

This model is trained to segment those parts which are infected by Covid-19 in CT-Scan.

• Trained with COVID-19 CT scans Kaggle data set.
• For the loss function, the combination ofweighted dice loss and surface losswas used and model trained with U-Net.
• Used Histogram Equalizationto make infection parts more visible and used augmentation to create more data for model.
• Sensitivity in test data set reached 0.934817 and AUC reached 0.933751.
• Tested with CT-MD Dicom data set and gets the perfect result as output.

• Link: GitHub
• Date: Oct 2020
• Service: TensorFlow, Keras, Pandas, OpenCV, Numpy

Developed the back-end side of a service that receives signals and tracks them to fetch specific features in each signal duration to have a better analysis of the cryptocurrency market. It is kind of a data mining project.
This service would pass the needed data from database to front-end using GraphQL.

• Client: Nurafarin
• Date: Sep 2020
• Technology: Node.js, GraphQL, and Mongoose ODM

This model would predict stock candles(whether it is a green or red candle) using 2D and 3D CNN architecture. It is my implementation of the idea of an article. This is a CNN-based framework, that can be applied on a collection of data from a variety of sources, including different markets, in order to extract features for predicting the future of those markets.

• Link: GitHub
• Date: Nov 2021
• Technology: TensorFlow, Keras, Pandas, Numpy

This project has consisted of two-part:

• the close prices of BTC in every minute candles from 24 hours ago needed to be gathered.

• the aforementioned part must be done for the next 24 hours in real-time.

• Link: GitHub
• Date: Feb 2021
• Technology: Python

Wrote a code that displays the status of 4 technical analysis method (RSI, RVI, Ichimoku, Bollinger Bands) in real-time with Python.

• Client: Nurafarin
• Date: Feb 2021
• Technology: Python, Ta-Lib, Torent, Pandas