Scribd
Upload
256 views

Algo Bot

The document describes an algorithmic trading bot that uses machine learning techniques like random forest classifiers to make trading decisions. It collects past cryptocurrency price data to build a training dataset and uses it to train a random forest model to predict future prices and determine optimal trading times and prices. The bot aims to make trading easier for users by automatically executing trades based on the model's predictions while also implementing risk management strategies like stop losses. Testing showed the bot was able to achieve profitable outcomes with reduced effort compared to manual trading. Future work could focus on improving the bot's efficiency and conducting longer live testing periods.

Uploaded by

Sumukha G C
Copyright
© © All Rights Reserved
Available Formats
Download as PPTX, PDF, TXT or read online on Scribd
256 views

Algo Bot

The document describes an algorithmic trading bot that uses machine learning techniques like random forest classifiers to make trading decisions. It collects past cryptocurrency price data to build a training dataset and uses it to train a random forest model to predict future prices and determine optimal trading times and prices. The bot aims to make trading easier for users by automatically executing trades based on the model's predictions while also implementing risk management strategies like stop losses. Testing showed the bot was able to achieve profitable outcomes with reduced effort compared to manual trading. Future work could focus on improving the bot's efficiency and conducting longer live testing periods.

Uploaded by

Sumukha G C
Copyright
© © All Rights Reserved
Available Formats
Download as PPTX, PDF, TXT or read online on Scribd
You are on page 1/ 20

ALGORITHMIC TRADING BOT

Under the guidance of


Dr. Megha P Arakeri

Submitted by
Sumukha G C – 1MS18IS112
  Dept of Information Science & Engineering
Seminar - ISE
INDEX
 Abstract and Motivation

 Problem Statement

 Objective

 Literature Review: The Architecture of the Proposed Solution,


Process, Flowchart
 The Deep Learning Technique Used

 Results and Conclusion

 Summary and Future Work

 References
MOTIVATION
 Industry reports suggest global algorithmic trading market size
is expected to grow from $11.1 bn in 2019 to $18.8 bn by 2024.

 Simple and efficient bot for use of common man.

 Most traders often have less time to learn to trade and to make a


successful profitable trade is a whole other game.
PROBLEM STATEMENT
 To help make life of a trader easy through the use of an
algorithmic trading bot that executes orders utilizing
mechanized pre-modified trading guidelines.

Reference : Algorithmic Trading Bot Medha Mathur, Satyam Mhadalekar, Sahil Mhatre,
Vanita Mane Department of Computer Engineering, Ramrao Adik Insitute of Technology,
Nerul, Navi Mumbai 
OBJECTIVES
 Creating a python bot that can execute trades at the best
possible prices and times guided by the following:
• The predictions made by a deep learning  technique.
• Financial Trading strategies, Market conditions, Entry and Exit
Points, Stop Loss value, and Strategy specific parameters like
MA etc.
 Constantly send instant emails to the user alerting about
the trading actions initiated. 
FLOWCHART
PROCESS
DATASET DESCRIPTION  

 Alpaca API and Binance is used to fetch past data and put it

into a dataset. The dataset comprises Date , Open Price , High


Price , Low Price , Close Price and Volume traded for that
particular Cryptocurrency day wise. 
TOOLS AND TECHNOLOGIES USED

Hardware
 Inter Core i7-9750H CPU
 2.60GHz

 8GB RAM

 TB HDD

Software
 Browser like Google Chrome, Mozilla Firefox, Microsoft Edge etc.
 Visual Studio Code Editor

 PgAdmin4 for PostgreSQL database

 Python 3.7

 Django Framework

 Alpaca Trading and Paper Trading Account


ALGORITHMIC TRADING
USING RANDOM FOREST
CLASSIFIER
DEFINITION

• It is based on ensemble technique


• Random Forest is a classifier/regressor that contains a number
of decision trees on various subsets of the given dataset and
takes the average to improve the predictive accuracy of that
dataset
PROCESS OVERVIEW
ALGORITHM
procedure RF(D)
f = new Array() --> create a new forest
for do i = 0 to N
Di = bootstrapSamples(D) --> randomly select n data from D
attributei = RandomAttributeSelection(Di)
DTi = new DecisionTree()
DTi.train(Di,attributei)
f.add(DTi)
end for loop
return forest f --> return the new forest
end procedure RF(D)
GRAPH OF OOB ERROR VS
NUMBER OF ESTIMATORS
RISK MANAGEMENT

• Trade Size
• Stop Loss
• Hedging
ACCURACY VS DIFFERENT MACHINE
LEARNING MODELS
RESULTS / CONCLUSION
 Algorithmic Trading Bot makes it easier for both new
traders as well as established ones in getting profitable
outcomes with minimized effort, time, and loss.
 The integration of Financial Knowledge with Machine
Learning is a demand of future Trading and enhances
both Performance and Revenue.
SUMMARY & FUTURE
WORK
 Algorithmic Trading Bot makes it easier for both new traders as
well as established ones in getting profitable outcomes with
minimized effort, time, and loss.
 The integration of Financial Knowledge with Machine Learning
is a demand of future Trading and enhances both Performance
and Revenue.
 Improve bot efficiency.

 Longer periods of live testing of the overall complete bot to


ensure best performance statistics.
REFERENCES
• https://arxiv.org/abs/1605.00003
• https://www.sciencedirect.com/science/article/abs/pii/
S0957417413009731
• https://pdfs.semanticscholar.org/
6dd9/0b152dcb5cda46e3640e3648e46d28838372.pdf
• Algorithmic Trading Bot Medha Mathur, Satyam Mhadalekar,
Sahil Mhatre, Vanita Mane Department of Computer
Engineering, Ramrao Adik Insitute of Technology, Nerul, Navi
Mumbai
Thankyou!

You might also like