MonPrix

Team Die Torch

Abstract

The stock market is one of the most enriching platforms for investment. It enables quick growth and success for companies. However, given the uncertainties and inevitable volatility, how can one trust the market cycle and go with it?

People have lost thousands of rupees in stock market downfalls in the past. Due to this, whenever the stock market crashes, people incline towards panic withdrawal which in turn leads to more losses. It a vicious circle. So, what’s the solution? Patience and time-tested strategies to prevent further loss of capital and deal with stock market fear.

Looking at the current market conditions, we thought that it was the right time to develop some solutions in which you can break this loop.

The Problem

A large population believes that investing in the stock market is risky and full of losses.
While in reality, the stock market encourages good decisions in a company and enables them to grow quickly. It also helps investors make money.
The simple reason for this myth is that beginners aren't educated enough and end up losing a large amount which discourages them.
Our model provides insights to the user, educating them about profitable investments in the stock market.

The Solution

Instead of throwing someone into a simulation or in between paragraphs of long marketing theory we offer simpler solutions
We embrace learning by letting users compete against our AI Bot.
After the user enters a time period and stock symbol, we run a simulation to show strategies that could be used to make a profit on this stock.
This breaks the myth for two reasons -
1. Establishes the Market isn't gambling
2. Making a profit based only on previous trends is easy enough for a simple bot to learn
Learning by competing this way is much faster and natural compared to reading theories about how others invested.

How are we different?

This is a unique approach to demystifying the stock market.

No sign ups required ( we don't steal data)
You learn by looking at where an AI bot traded money only by going through the charts.
We encourage you to use your own judgement to trade instead of recommendation.

Features:

Trading Bot
- Embrace learning by letting users compete against our AI Bot.
Real Time Stock Data
- Collects and analyses Stock Market's Data in Real Time Using Yahoo Finance
Real Time Sentiment Analysis via Twitter
- To Analyse and perform joint Sentiment analysis and Price Predictions for Stock Market's Data in Real Time Using Twitter API For AI Bot
Graphical Interface
- Analyses The Bot's Performance to let user know about Stock's Performance
Stock Prediction
- Predicting stocks for the next day.
Crypto Price Prediction
- Predicting crypto currency prices for the next 5 days.
Flowchart:

Setup:

You Need to Set-up an Virtual environment by installing dependencies mentioned in requirements.txt First.
```
1. start ML/app.py

2.Open build/index.html
```

Link to Presentation : Click Here

Tech Stack:

ML :

Deep Learning
LSTMs
Recurrent Neural Networks
Natural Language Processing
NLTK
Scikit-Learn
Tensorflow
Numpy
Pandas

Frontend Tech Stack :

Html and CSS
ReactJS

Backend Tech Stack :

NodeJS
ReactJS which is a NodeJS Framework
Flask which is a Python Framework

Database Tech Stack:

MongoDB

Demo :

Challenges we ran into:

Since our project requires a huge database of all the Stocks, it was initially a challenge to think of ideas to maintain and update the databases. For this, we can use any cloud-based platform.
The project requires a solid back-end with the integration of front-end using ReactJS Framework. Therefore being ML Developers, its implementation was quite a challenge as it was a new concept and we had to learn and implement a lot of new concepts in a very limited period of time.
But all in all, it was great fun building this project together.