Improve existing Reinforcement Learning Model in Python and Connect to Ninjatrader (code will be provided)
Bütçe $750-1500 USD
*Please only bid if you have experience with implementing Trading Algorithms*
The current model works from training the agent to placing live trades in Interactive Brokers (IB) through IBridgePy, with following attributes:
Actions: Buy, Sell, Hold
States: Long, Short, Flat
A State can transition in the following ways:
Flat to Long/Short by taking Buy/Sell Actions respectively, by Buying/Selling at market
Long/Short to Flat by taking Sell/Buy Actions respectively by Selling/Buying at market
At State transition, a reward is returned
State does not transition or the subsequent actions are same:
Long/short state does not change if action is Buy/Sell respectively or Hold
Enhancements:
1.) Ensemble:
Setup the model as ensemble learning model
[login to view URL]
2.) Session Time/Economic Calendar:
Trades will be taken during defined session times(RTH, ETH)
Economic calendar will be checked prior to entering a trade or going flat ahead of economic event
3.) Swing Calculation:
Prior to the start of each session, calculate the most occurring swing size (mode) on a given timeframe for the previous n number of sessions. This will be implemented using zig-zag logic
4.) Add Target, Stop-Loss and Break Even(BE) Logic
Target/SL/BE logic will be implemented
Value for Target/SL will be determined by swing calculation
Additional trade metrics will be calculated (time in trade and drawdown) and will be passed to reward function for calculation
5.) Feature Engineering
Custom indicators will be created based on zig-zag to mark wick on swing high/low candles and fed as a feature to the Model. When price re-visit the previous swing high/low, the indicator will display 1 otherwise 0
Create features based on instrument other than trading instrument (Trading ES, create additional features for NQ, YM, Bonds, etc)
6.) NinjaTrader Integration
Place orders with OCO
Able to run the in re-play mode for back-testing
7.) Program Structure:
Setup modular structure; model can be separated from Ninjatrader and configured with other system
Instrument file with information on tick value, size, etc
Bu iş için 14 freelancer ortalamada $1218 teklif veriyor
hi...i m experienced in machine learning and implementing reinforcement learning....i would love to work on this interesting project with you. lets discuss in detail....thanks
Hi!, I am a professional data scientist with 5 years of experience. I hold an MBA and first Degree in statistics which provides me with the necessary background to handle your project. I've carefully checked your requ Daha Fazla
Good day prospective client. I am a senior software developer who is majorly focused on creating trading software/bot, some of which include: - MetaTrader 4,5 Expert Advisors and Indicators - Tradingview Pine script St Daha Fazla
Thanks for your posting. I am a senior machine learning/backend engineer. I would like to help you with the following experiences. - Developed the trading bot based on customer-specified trading rules(simple buy/sell Daha Fazla
Hi, I hope you are doing fine. I have almost 10 years of experience in machine learning algorithms. I can implement various types of artificial intelligence algorithms including yours with Matlab, Python and etc. I hav Daha Fazla
Hello, I am Manana from Armenia Yerevan. I read your description in detail. I have experiences at web security and can help you. I am very excited on your project, and I am ready to start work immediately. I have skil Daha Fazla
I am Qaiser. I have a strong grip on Data science ML, AI I am available 24/hr for clients. 23 projects already in AI, ML Data science with 100% perfect. come in chat and enjoy my services
Hi, We are probrAIn and we are based out of Toronto, Canada. probrAIn is founded by 3 members with a collective experience of 10 years in corporate settings with experiences ranging from data science/analytics, softwa Daha Fazla
We have more than 10 years experience doing projects in Machine learning, Reinforcement learning and also payment network. In addition we developed some projects of reinforcement learning in C# and python. So we Daha Fazla
Hello, Thanks for posting, I read your description carefully. We already working on 2 reinforcement learning projects, one is AWS deepRacer and other is using RL to decide over delivery and under delivery of ad campa Daha Fazla
Hello, Sir. I read your opinion seriously. Working experience with [login to view URL] api for tradding. please share I can help here. I have experience of 7 years and done similar projects in done similar projects before. Daha Fazla
Algorithmic trader with over 10 years of experience and over 30 doubled accounts, Ph.D. from University of Alberta in Computational Mathematics + Ms. Degree from MIPT. Leading small but highly seasoned team of devel Daha Fazla
