Majid's Projects
Comparison of various machine learning models on a trading algorithm with the goal of identifying a model capable of producing returns across various stocks.
Program used to filter through a list of loans and generate a csv of the most affordable options.
This project is an analysis of peer-to-peer lending data using Logistic Regression and oversampling to identify whether or not oversampling helps the model better predict healthy and risky loans.
This project uses deep neural networks to create models that predict whether or not a business will become a success if invested in.
This project uses machine learning models to train and evaluate a trade algorithm with the goal of increasing the cumulative returns relative to the actual returns.
Blockchain-based ledger system built in Python with a user-friendly web interface.
Application that allows user to select a candidate, calculate the wage to hire that candidate, and send the transaction from their address to the candidates via an easy to use web interface.
Joint savings account smart contract created with Solidity.
ERC-20 token, crowdsale, and deployer contracts to crowdsale a new token called KaseiCoin (KSC) and allow users to purchase newly minted tokens.
Program that intakes bank and user data and generates a csv of available loans from select banks.
Financial analysis of Bitcoin to identify arbitrage opportunities using historical data.
Analysis of fund portfolios and the S&P 500 to identify ideal portfolios for a firm's suite of fund offerings.
Two financial analysis tools to help credit union members 1. visualize current savings accounts and determine if enough funds are available to create an emergency fund and 2. assist in financial planning for retirement
Analysis of sale price per square foot, gross rent, and available housing units for various neighborhoods in San Francisco.
Analyze single ETF assets and ETF portfolios using python, pandas, and SQL. Deployable as a web app using voila.
Analysis of different cryptocurrencies using unsupervised learning to cluster based on performance in different time periods.
Analysis of stock prices, Google trends data, and sales data to identify seasonal trends and forecasting using the Prophet library.
Analysis of Bitcoin and the S&P 500 to identify best investment opportunity. Analysis includes long-term holding, shorter-term trading, and Monte Carlo simulations to find the most profitable opportunities and potential future investments.
CNN that classifies fresh and rotten fruit. Transfer learning using the VGG16 model and data provided by Kaggle.
Config files for my GitHub profile.
Simple program using user input data to generate a QR code.
Simple web application using the Prophet library and financial APIs to provide stock price forecasts.
New token creation and ICO