Abstarct: General descxription: Determining the price is one of the most challenging issues in financial markets. Although the financial market is intentionally designed to be almost unpredictable, it has been proven that the price change trends in these markets can be predicted. The present thesis aims to find the trend of the prices in the market. To accomplish this goal, which literally means determining the upward, downward or flat movement of the market, proposed hybrid automated algorithms would be used. Method descxription: This research aims to propose an automated method using a combination of some methods to investigate price movement. In effect, fundamental algorithms are baxsed on Kinetic formula either in combination with an indicator, or with deep reinforcement learning, all of which are in order to make automated trading algorithms. Due to the fact that using the aforementioned methods separately would produce satisfying results, their combination probably led to it too. In addition, we tried to make the most of each method, in their combination. Therefore, in one word, our proposed method is an integration of Kinetic energy and some models, such as deep reinforcement learning models. Findings: For verifying the proposed models, some experiments have been done on a data set from actual financial data, which are provided by Forex. baxsed on the experiments, it is revealed that the proposed method could outperform the DL model, considering several metrics such as Recall, Precision, Accuracy, and F-measure. The proposed methods could achieve better results baxsed on the result. To be more precise, the proposed method as a hybrid model baxsed on the combination of Kinetic formulas and RSI, called RSIK, could outweigh other approaches on GBPUSD dataset, with approximately 95% average precision on prediction. On the other hand, for the second approach gaining more total profit is determined for the evaluation section. In fact, the profitability percentage of proposed method, DRLK, is 57.14% on the APPLE company dataset (APPL), while the method of comparing article, is 54.55% considering the same dataset. The use of the proposed method with the hierarchical reinforcement learning approach also had a significant impact on improving the accuracy of the prediction results.