Development of Trading Bot for Stock Prediction Using Evolution Strategy

EasyChair Preprint 6739

Abstract

This paper presents the idea of implementing a trading bot that can implement trading strategies in an automated fashion using minimal or limited interactions with the user. This Bot has been trained explicitly to understand the trading practices and the types of orders, the quantity, the profit margins and when to exit the trade. One of the most important objectives of the trading bot is to function according to the user preferences i.e., to produce stream quotes, in order to get price information which would further help in calculating indicators, placing orders and organizing the user data. The bot is also capable of checking any fluctuations in the market, study the data and cautiously contribute on the stocks that give at least 1 percent return using algorithms based on sentimental analysis and deep evolutionary strategies. The data from Twitter has been provided to the bot to make it understand and asses the market better. Using the current strategies on virtual outcome, simulation outlines that the bot used in the project is more proficient than the existing ones that are based on other different machine learning algorithms.

Keyphrases: Artificial Intelligence, Day Trading, Deep Evolution Strategies, Evolution Strategy, Machine Learning Algorithm, Reinforcement Learning, Sentiment Analysis, Sentimental Analysis, Stock Prediction, Trading Bot, machine learning, stock market

@booklet{EasyChair:6739,
  author    = {Ashish Bali and Archit Madan and Aayush Upadhyay and Piyush Sah and Vibha Nehra},
  title     = {Development of Trading Bot for Stock Prediction Using Evolution Strategy},
  howpublished = {EasyChair Preprint 6739},
  year      = {EasyChair, 2021}}