Abstarct: This study presents a method for performing optimal algorithmic transactions using machine learning methods. Here, we do not use ML forecasting methods as the primary trading tool because they are usually not efficient and applicable in real conditions, so we have tried to use these methods as an expert consultant to manage the risk of trading strategies. The Forex market is a suitable option for investing and conducting experiments in this field due to its importance and stability compared to other financial markets. Therefore, we took advantage of financial Forex time series data to train a deep learning-baxsed model that can classify trading strategy production signals with profitable and unprofitable labels. We used the ISTM network to design the risk management model because it was more optimal, more accurate, and had fewer reported error values compared to other appropriate methods in forecasting financial time series. In line with the training and productivity of the proposed risk management model, two trading strategies were designed and proposed baxsed on statistical methods and technical indicators, which have a good potential for making profitable transactions. The first strategy (S1) generates a signal to enter the transaction using calculations baxsed on a stochastic oscillator. To close it, it uses the determined stop-loss and the resistance/support levels of the Donchan channels and has a trading structure suitable for fluctuating conditions and sideway markets. The second strategy (S2) uses the directional average index and Bollinger bands to generate signals. It uses the middle band of Bollinger bands and the predetermined stop-loss to close trades, and its structure is more appropriate to the markets with trends. By observing the system's performance, we find that the trading results baxsed on the signals generated by the strategy equipped with the risk management tool, compared to the time that does not use this tool, have improved impressively. Profit growth from transactions using risk management tools is always positive and the lowest reported profit growth is 15% in transactions. In contrast, the percentage of profit growth in transactions has been reported up to several tens of percent