Effective Crude Oil Trading Techniques Using Long Short-Term Memory and Convolution Neural Networks Wisaroot Lertthaweedech, Pittipol Kantavat, and Boonserm Kijsirikul Department of Computer Engineering, Faculty of Engineering, Chulalongkorn University, Bangkok, Thailand Email: 6370267221@student.chula.ac.th, {pittipol.k, boonserm.k}@chula.ac.th Abstract—Crude oil plays a vital role in the global economy and forecasting crude oil prices is crucial for both government and private sectors. However, the crude oil price is high volatility, influenced by various factors and challenging to predict. Thus, various machine learning techniques have been proposed to predict crude oil prices for decades. In this study, we propose an Artificial Neural Network (ANN) with different combinations of Convolutional Neural Networks (CNN) and Long Short- Term Memory (LSTM) to improve the trend forecasting of crude oil prices for better trading signals compared to traditional strategies. As the crude oil price is a time series data, it is appropriate to apply CNN and LSTM for forecasting. The concept of our model is that CNN could detect features or patterns in different locations of time series data, while LSTM could maintain both short-term and long-term memory along with time series data. The collaboration of their abilities could help the neural network model understand complex relationships of historical data and trends of crude oil prices. Our study found that the combination of CNN and LSTM could significantly enhance trading performance in the long run. Index Terms—crude oil trading, machine learning, deep learning, trading signal, technical analysis, artificial intelligent I. INTRODUCTION Crude oil is a commodity that significantly impacts the world economy because 30% of the overall energy supply in the world uses crude oil as the source of energy [1] Products of refined crude oil are used in various economic activities such as power generation, raw material for the petrochemical industry, and transportation by vehicles, ships, and airplanes. Therefore, crude oil price possesses direct influences on many industries that are related to these economic activities. In addition, the crude oil price is a vital factor in the global economy for inflation forecast, monetary policy, and fiscal policy by the government sector [2]. However, the crude oil price is very volatile. It depends on the dynamic condition of demand and supply, including the growth of economic activities, technology, alternative energy like accepted May 26, 2022. natural gas, coal, renewable energy, black-swan event like the coronavirus disease of 2019 (COVID-19). There are several research on crude oil price prediction models for decades, such as Support Vector Machine (SVM) [3], [4], Autoregressive Integrated Moving Average (ARIMA) [3]-[5], Random Walk [3], Genetic Algorithm (GA) [6], Generalized Autoregressive Conditional Heteroskedasticity (GARCH) [7], Vector Autoregressive (VAR) [8], Error Correction Model (ECM) [7], [9], Thanks to the development of computing technology, machine learning models with complex algorithms, such as deep learning, are more popular and have performed higher performance in recent years [10]. Different types of deep learning models were proposed to implement various asset price forecasting models [11]. Convolutional Neural Network (CNN) was one of the most popular types of Artificial Neural Network (ANN) layer for stock price prediction. CNN could learn for feature selection automatically [12]. Gudelek et al. [13] used 2D CNN to classify types of stock price movement from the history of stock price and technical indicator data. They proposed 2-layers of CNN followed using a 3×3 filter size for both layers. The number of filters was 32 and 64 for each CNN layer, respectively. The performance evaluation results indicated that their proposed CNN models could predict stock prices movement with high accuracy and outperform Buy & Hold strategy. Tsantekidis et al. [14] proposed multi- layers of CNN to predict stock price movement using high-frequency time series derived from the order book. Their CNNs model was composed of 2 sets of convolution and pooling layers followed by two dense layers. They compared model performance with the linear SVM model and MLP model. They found that the proposed CNN model could predict stock price with higher accuracy than other models. Lee et al. [15] applied CNN with Deep Q-Network to predict various stock prices and perform the trading test in many stock markets globally. The proposed model utilized stock chart images as input for CNN as a function approximator. Then, CNN created feature maps as a representation for action. The portfolio from their backtest performed well in many stocks market over 12 years of testing. The Long-Short Term Memory model (LSTM) was another popular type of ANN layer for stock price Journal of Advances in Information Technology Vol. 13, No. 6, December 2022 645 doi: 10.12720/jait.13.6.645-651 Manuscript received March 10, 2022; revised May 17, 2022;