Machine Learning

Developer needed for creating a Deep Neural Network to train using my data sets My data sets are created using Renko Bars for ES and SPY data from 1997. Each Renko Bar generates 154 (columns) Numerical values. Data sets are created using Renko bars of various sizes. FOR ES and SPY data is from 1997. Training to decide entry and exit price for security (SPY & ES): buy low and sell high, sell high and buy low, profit/draw-down ratio > 5 Training to decide entry and exit price for options: Buy low and sell high only. No shorting. There are 39 Renko bars brick sizes for es and spy. For Options, data Brick sizes are 6 to 13 Familiarity RDMM (REINFORCED DEEP MARKOV MODELS WITH APPLICATIONS IN AUTOMATIC TRADING). 2011.0439 PDF is enclosed Reinforcement and mean-field games in algorithmic trading - Sebastian Jaimungal [login to view URL] Use Sharpe ratio to optimize for each length of tick charts Then combine the optimal from each tick chart and combine with all the tick charts to find the optimal entry and exit During the market (US) hours combine the ES, SPY, and Options data also in deciding entry and exit Options on Spy are calls and Puts. Calls brick sizes are 1 tick to 6 ticks. Puts brick sizes are 1 to 13 ticks. Data is generally no longer than 180 days for near-the-money options. Puts deep in the money are for a few years only. Options data are used for Buy low and sell high only, No sell high buy low for Calls and puts. If the market, is in uptrend calls are the play If the market is in downtrend puts are in the play REINFORCED DEEP MARKOV MODELS WITH APPLICATIONS IN AUTOMATIC TRADING https://export.arxiv.org/pdf/2011.04391.pdf#view=fitH&toolbar=1 This program is written for high-frequency trading. This logic has to be modified to fit the Renko Bar charting. I have enclosed a few datasets for es, spy, call and put off the "SPY". Exclude the first five hundred Renko bars data from the learning algorithm.