Stock Grading System

Summary The goal of the project is to build a web-based platform using Machine learning and deep learning and financial techniques to predict the future behavior of the stocks in the Nasdaq stock market. Analyzing the behavior of the stocks in the Nasdaq stock market is a cha

Project Title

Stock Grading System

Project Area of Specialization

Artificial Intelligence

Project Summary

Summary

The goal of the project is to build a web-based platform using Machine learning and deep learning and financial techniques to predict the future behavior of the stocks in the Nasdaq stock market. Analyzing the behavior of the stocks in the Nasdaq stock market is a challenging task due to the high level of uncertainty, investors sentiment, and many other factors. There are mainly two types of analysis for analyzing stock behavior, fundamental analysis, and technical analysis. Technical analysis involves the analysis of historical prices of the stock along with the trading volume. Fundamental analysis involves the analysis of the intrinsic value of the stock by considering the different economical and financial factors. This makes investment in the right stock at the right time cumbersome for most of investors. Especially new investors find it difficult to do fundamental as well as technical analysis of the stock which is necessary before adding any stock to the portfolio. Considering this fact, the primary goal of developing this platform is to make fundamental and technical analysis of the stock easy for a better investment decision. Moreover, we will also include sentiment analysis of the stock by taking the sentiment from social media platforms like Reddit, and Stock Twits to further enhance the credibility of the results. Thus by involving fundamental analysis, technical analysis, and sentiment analysis, we will give a particular score to the stock which will enable the investors to make a better investment decision.

Project Objectives

Aim and Objectives:

This stock grading system will be a web-based platform that will provide all the information necessary for getting intuition about the stock for investment at the same place so that that investor will be saved from the painful effort of manual financial calculations. This is the primary aim of building this project.

The objectives for building this system are

• Fundamental analysis of the stock by using different financial ratios related to stocks.
• Technical analysis of the stock via using the past historical data of stock prices as well as trading volume.
• Sentiment analysis of the stock by getting the sentiment from social media platforms. e.g., Reddit, Stock Twits

Project Implementation Method

Implementation Details:

Stock Grading System will be a web-based machine learning project in which we will deploy Price prediction and sentiment analysis models. For the price prediction, we will use the Long Short Term Memory Network which is the architecture of the Recurrent Neural Network. For the Dataset, we will use the Yahoo finance website. In the production, the model will get the live data from the Yahoo finance website in order to make the future prediction of the stock.

For the sentiment analysis model, we will use FinBERT which is the pre-trained model available at Huggingface, a famous platform for solving the Natural Language processing. We will finetune the FinBERT model using the dataset available at Kaggle. Moreover, as the deep learning models require a large amount of data, we will make our own labeled dataset for fine-tuning of the FinBERT model.

Moreover, we will use paid APIs for Twitter, Reddit, and StockTwits to get the discussions of the people related to the stock being searched by the user at run time. We will calculate the sentiment score based on the model output of different comments of the users on the social media platforms. This sentiment score will eventually help us in calculating the overall grade of the stock.

For Grading, we will use the domain-specific(Finance) knowledge to provide the overall grade of the stock based on many financial ratios and indicators.

Evaluation Parameters

For the fundamental analysis, we will evaluate the score by comparing it with the different similar websites. Though every website has its own strategy, but it will give us the idea of whether we are scoring the stock in a right or wrong way. Moreover, we will also get the assistance of a finance specialist.

For technical analysis, as we are involving the machine learning model for this problem, we will evaluate the prediction results using the Means Squared Error (MSE), Root Mean Squared Error (RMSE), Mean Absolute Error (MAE), R Square, and Adjusted R Square.

For Sentiment analysis, we will select 100 latest posts and check the sentiment manually to analyze whether our model is giving us accurate results or not.

Development Process:

We are using the incremental and iterative development process in this project. The reason for adopting this methodology is that we need to work step by step to meet the requirements. Also, this method suits the event of changed requirements. This method ensures that the developers would make changes early during the process rather than waiting till the end. This will ultimately save both time and the budget used for the resources. The iterative development process ensures that the improvements are made on an ongoing basis and helps in delivering the product on time and of high quality.

Benefits of the Project

Benefits:

• The platform will help investors who want to get an idea about the current position of the stock.
• Analysts will get help from the grading system to analyze the stock.
• This platform will be effective for financial advisors who want to help their clients
• Financial Writers who want to get insights about the current behavior of the stock
• The platform will help stockbrokers in buy and sell decisions on the behalf of their investor Clients
• The platform will be beneficial for finance students.

Technical Details of Final Deliverable

Deliverables

• Price Prediction Model

This Machine learning or deep learning-based model will predict the short-term price of U.S stocks.

• Sentiment analysis model

This deep learning-based model will analyze the social sentiment of Twitter, Reddit, and Stock Twits of any stock which users want to analyze. The final output will be the sentiment score from -100 to +100 which will give the idea to the user whether the stock is bullish or bearish in the U.S stock market.

• Grading Strategy

This will cover the fundamental analysis of the stock as it will include many financial ratios based on which the system will give a particular grade to the stock.

• Product Prototype

The prototype will consist of the frontend as well as backend. For the frontend, we will use Javascript and its framework Next.Js. For the backend, we will use the Python web framework Django along with MySQL database.

• Project Thesis

Complete detail of the project from implementation to production.

Final Deliverable of the Project

Software System

Core Industry

Finance

Other Industries

Core Technology

Artificial Intelligence(AI)

Other Technologies

Sustainable Development Goals

Quality Education, Decent Work and Economic Growth, Industry, Innovation and Infrastructure

Required Resources

If you need this project, please contact me on contact@adikhanofficial.com