High-Speed Hardware System for Stock Market Prediction Using Machine learning.

Stock market prediction is one of the most trending problem in today?s world and the only aim of stock market prediction is to make profit and reduce capital risks. According to the efficient market hypothesis theory, it?s almost impossible to predict the stock market prices with 100% accuracy as it

Project Title

High-Speed Hardware System for Stock Market Prediction Using Machine learning.

Project Area of Specialization

Artificial Intelligence

Project Summary

Stock market prediction is one of the most trending problem in today’s world and the only aim of stock market prediction is to make profit and reduce capital risks. According to the efficient market hypothesis theory, it’s almost impossible to predict the stock market prices with 100% accuracy as it takes into account many factors like the political upheaval, interest rates, current events, exchange rate fluctuations, natural calamities and much more.

The invention of machine learning techniques makes it possible to predict the stock market direction with maximum accuracy and increase profit and reduce capital risks by reacting with high speed to those predictions using High-speed hardware like FPGA (Field Programmable Gate Array). High-Speed Hardware System promises an increase in profit of about 0.0005 per share or a twentieth of a penny.

This project aims to design a High-Speed Hardware System to predict stock market prices with maximum accuracy using machine learning techniques. This project will be mainly divided into three parts, in the first stage we will be developing an algorithm in python for stock market prediction using machine learning technique in which first of all we will train our ML model on Stock market data and after analyzing the capability of our developed model in software, in the second stage we will implement our algorithm on FPGA (High-Speed Hardware) using VHDL (Verilog High Description Language ) to achieve high speed to maximize profits and reduce risks and finally in the third stage we will require to capture real-time data for the stock market so we will interfacing FPGA with Ethernet  (10 gigabytes)  for real-time stock data.

                                                                  Figure 1: Project Different Stages Overview

The Final Product out of this project will be like a real-time stock data will be streaming through Ethernet into the FPGA Hardware system, algorithm implemented inside the FPGA will make a decision based on the stock data like buying/selling of stock and after that those decisions will be sent back to the stock market.

                                       

                                                                     Figure 2: Final Product Overview

Project Objectives

Pakistan is a developing country that came into existence almost 75 years ago and for any developing country financial market system is one of the most important thing in its progress but unluckily we still utilize the manual traditional software-based methods for the financial market management system which are too slow as well as manual.

The main objectives of this project are

• To Provide Stock Market prediction with maximum accuracy using machine learning.
• To increase profit of the investors in the stock market using High-Speed Hardware (FPGA).
• To Provide a scalable, cheap, and fast stock market capital management hardware system.
• To Provide a system that is automatic and intelligent with less or no human interface.
• To Provide a system that reduce capital risks of investors during the market crash.
• To Provide a system that works on every market like cryptocurrencies, forex and stock market.
• To Provide a system that can scan multiple markets and exchanges. It enables traders and investors to find more trading opportunities, including arbitraging slight price differences for the same asset as traded on different exchanges.

Project Implementation Method

The Implementation of this project will be done in following parts:

Designing and implementation

The designing and implementation phase of this project is started in mid of October and we will continue working on the following aspects. 

Designing and implementing stock market data extractor

We will design and implement a python code to extract free stock data from broker or stock market. The data we extract will include High, Low, Close, Open and Volume data columns.

Designing and implementing stock market data organizer

We will design and implement a python code on extracted stock data to make it organize, clean and readable for humans.

Designing and implementing an algorithm for stock market prediction using machine learning technique

We will design a python code to develop an algorithm for stock market prediction using machine learning model. This designed algorithm will be the main brain of the system which will be responsible for controlling and generating all the financial decisions in the financial management system and then we will implement this algorithm on stock market data.

Designing Back tester for developed algorithm

We will design a back tester in python to analyze and check the intelligence, accuracy and decisions making of developed algorithm on financial stock market systems.

Designing a hardware-based code for developed algorithm

After designing an algorithm in software (python) successfully now it will be a perfect time to develop a hardware-based code for an already developed algorithm using VHDL.

Implementing VHDL algorithm on FPGA board (High-Speed Hardware)

We will implement already designed VHDL-based algorithm on the FPGA board. We will test our VHDL-based algorithm on dummy data in FPGA board.

Designing Ethernet interface for FPGA

After verifying the intelligence, accuracy and decisions making of VHDL based algorithm, we will design an interface to connect Ethernet to the FPGA using VIVADO software for extracting the real-time data.

Designing an API for Full-duplex communication using Ethernet

After interfacing Ethernet with FPGA board properly we will design an API for Full-duplex communication between stock market and our FPGA board using C coding, so data can be transmitted as well as received through FPGA board.

Designing an API for Stock market real-time data parser

After receiving stock market data in real-time from stock market we will design an API for Stock market data to parse the data and make it processible for FPGA algorithm.

Designing a full data path from Ethernet interface to FPGA algorithm

This will be the last step in designing and implementation. In this part, we will design a full two-way data path from the Ethernet interface to FPGA algorithm and vice versa, so that the stock data can be received and predictions on that data can be transmitted simultaneously.                       

      

                                                                             Figure 3: Project Implementation Overview

                                                                                   

Benefits of the Project

There are many benefits of this project for Pakistan financial and stock prediction systems as well as for local investors Some of the benefits are follow as below:

• To Provide Stock Market prediction with maximum accuracy using machine learning for local Investors.
• To increase profit of the investors in stock market using High-Speed Hardware (FPGA). It promises an increase in profit of about 0.0005 per share or a twentieth of a penny.
• A system that can be used for High-frequency trading as well by hedge funds in Pakistan.
• A scalable, cheap, and fast stock market capital management hardware system.
• A system that is automatic and intelligent with less or no human interface.
• A system that reduces capital risks of investors during a market crash.
• A system that works on every market like cryptocurrencies, forex and stock market.

A system that can scan multiple markets and exchanges. It enables traders and investors to find more trading opportunities, including arbitraging slight price differences for the same asset as traded on different.

Technical Details of Final Deliverable

The final deliverables are divided into Hardware and Software components:

Software Deliverables

In our project Software Deliverables will be:

• A Python code that will be able to provide and organize Stock market data with one-time execution.
• A python code that will make the stock data into readable format.
• A Python code that will be able to provide predictions about the stock market direction in real-time for the public and investors.
• A Python code for developers and traders who want to check the accuracy of their developed systems on the stock market.
• A Converted Code from python to VHDL for high-speed integration.
• Ethernet Interfacing design using VIVADO software will be available online for public.
• API in C coding for two-way communication between transmitter and receiver.
• API for Stock market real-time data parser in VIVADO Software.

Hardware Deliverables

A fully developed High-Speed Hardware System for Stock Market Prediction Using Machine learning on the FPGA board that will make the financial system implemented in Pakistan more intelligent, robust, automatic, fast and accurate. This product will be highly desirable to the financial industry as well as Hedge funds in Pakistan to accurately predict the market with High Speed and Maximum Accuracy.

Final Deliverable of the Project

HW/SW integrated system

Core Industry

Finance

Other Industries

Others

Core Technology

Artificial Intelligence(AI)

Other Technologies

Others

Sustainable Development Goals

Decent Work and Economic Growth

Required Resources

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