Markov Assistant

Supports

• Short-Term stock price's up & down prediction
• Long-Term stock price's movement
• Trading Bot Backtesting
• Stop Loss Monitoring

Program Consists of:

Main Window

Portfolio

Register your interested stock, and manage with data that program provides.

Short-Term Stock Price Up & Down

Using Densed RNN algorithms, based on log-return profit, you can seize tomorrow stock prices' up & down.

Long-Term Stock Price Flow Prediction

Using DNN algorithm, its flow chart infected many other indexes, stocks.
You can thrive you portfolio more profitable and be acknowledged about prices a few months later.

Backtesting Your own algorithm

Your main trading algorithm consisted of different agent that works interdependent. Furthermore backtest your RL-model with specific days, you can measure your model works well.

Train New Backtest Model

• Symbol : what stock you want to train.
• Lags : previous days referenced by the model.
• Episodes : model's training cycle.
• Train Days : previous days that model will train.
• Interval : update interval of stock price.

Train New Stop loss Model

Trade Window

• Holding : Your one of your agent is for prevent trading while sideway market.
• Trading : Other agents, interact each other, trade a stock.

1 minute interval candle stick graph, bollinger bands supported.

Installing Environment

• CUDA: 12.4
• Python: 3.9.12
• Keras: 2.15.0
• numpy: 1.26.3
• pandas: 1.4.2
• PyQt5: 5.15.10
• pyqtgraph: 0.13.3
• tensorflow: 2.15.0
• yfinance: 0.2.36