Minor Project

You might also like

Download as pdf or txt
Download as pdf or txt
You are on page 1of 21

Stock Price Predictor

Done by Group 11 -
Ayush 18103076
Priyanka 18103095
Samridhi 18103108
Harman 18103128
Introduction

● General stock investors investing in stock market usually are not certain
about the stock market behavior.

● They face the problem of trading as they do not properly understand


which stocks to buy or which stocks to sell in order to get more profit.

● Analyzing all the stock market related information individually or


manually is tremendously difficult.
● As such, automation of the process is required. This is where Data
mining techniques help.

● Understanding that analysis of numerical time series gives close


results, intelligent investors use machine learning techniques in
predicting the stock market behavior.

● This will allow the investors to foresee the behavior of the stock that
they are interested in and thus act accordingly.
Purpose
● The project target is to create a Program that analyses previous stock
data of companies and implement these values in Deep Learning
algorithm to determine the value that particular stock will have in near
future with suitable accuracy.
● These predicted and analysed data can be observed by individual to
know the financial status of companies and their comparisons.
● Active Investors can then buy or sell the stocks according to the stocks’
future trends.

Technologies Used
1. Data Set of various Stocks over 10 Years of Time. (Using YahooFinance
API)
2. Deep Learning (LSTM) (For prediction)
3. Python (To implement ML algorithms and read data)
4. JAVA (For GUI)
Stocks Considered
● EXPE (Expedia Group, Inc)
● MMM (3M company)
● XOM (Exxon Mobil Corporation)
● BA (Boeing Company)
● DAI.DE (DAIMLER AG)
● GE (General Electric Company)
● HOG (Harley-Davidson, Inc)
● ITC.NS (ITC Ltd)
JavaFX
FRONT END

● JavaFX- JavaFX is a software platform for creating and delivering desktop


applications, as well as rich Internet applications that can run across a wide variety
of devices.
● CSS- Cascading Style Sheets has been used for styling the application.
● The project uses JavaFX for the graphical user interface.
● ML model produces 3 graphs per company and a csv file to store the present
day’s and the next week’s data.
● Java is then used to retrieve and display these 3 images and the data for every
company using JavaFX.
● The main page contains the names of the companies and once the user selects
the company, the name is transferred to the back end and the data displayed
accordingly.
Features
● Comparison of actual and predicted data
A detailed analysis of the Stock market is presented to the user. The analysis
contains the performance of the listed companies for a certain interval of days v/s the
predicted values by the model.
● Value of the stock for the next 7 days in the form of a graph
The system provides the predicted value for the next 7 days in the form of a graph for
the convenience of the user.
● Graphical analysis of the stock one month later
While the ‘next 7 days’ graph is helpful in providing the exact predicted value of the
stock, the one month graphical analysis provides a trend of the stock for the
qualitative analysis by the user.
Design - UML
Diagrams ❖ Class Diagram
❖ Use Case Diagram
❖ Activity Diagram
❖ Sequence Diagram
❖ State Diagram
Class
Diagram
Use Case
Diagram
Activity Diagram 1
View Trade Exchange / View
Predicted Outcome
Activity Diagram 2
Compute Result and
Performance
Sequence
Diagram 1
View Trade Exchange / View
Predicted Outcome
Sequence
Diagram 2
Compute Result
and Performance
State Diagram
Recurrent Neural
The Problem of Long-Term
Networks Dependencies
They are networks with loops in them, allowing
Sometimes, we only need to look at recent
information to persist, so that the machine doesn’t
have to start from the scratch. information to perform the present task. But
there are also cases where we need more
context.

Unfortunately, as that gap grows, RNNs


become unable to learn to connect the
information.
Short Term Dependency Long Term Dependency

(RNNs fail)
LSTM Networks LSTM has the ability to add or remove
information from the cell state using gates.
(Long Short Term Memory)
Its architecture comprises the cell, input gate,
output gate and forget gate
LSTM networks are an extension of recurrent
neural networks (RNNs) mainly introduced to ● The cell of the model is responsible for
handle situations where RNNs fail. keeping track of the dependencies
For example: between the elements in the input
● RNNs fail to store information for a longer sequence.
period of time ● Input gate controls the extent to which a
● No finer control over which part of the new value flows into the cell
context needs to be carried forward and ● Forget gate controls the extent to which
how much of the past needs to be a value remains in the cell
‘forgotten’ ● Output gate controls the extent to which
the value in the cell is used to compute
the output activation of the LSTM unit.
Thank You

You might also like