Forecasting Commodity Prices Using Long Short Term Memory Neural Networks

This paper applies a recurrent neural network (RNN) method to forecast cotton and oil prices. We show how these new tools from machine learning, particularly Long-Short Term Memory (LSTM) models, complement traditional methods. Our results show that machine learning methods fit reasonably well with the data but do not outperform systematically classical methods such as Autoregressive Integrated Moving Average (ARIMA) or the naïve models in terms of out of sample forecasts. However, averaging the forecasts from the two type of models provide better results compared to either method. Compared to the ARIMA and the LSTM, the Root Mean Squared Error (RMSE) of the average forecast was 0.21 and 21.49 percent lower, respectively, for cotton. For oil, the forecast averaging does not provide improvements in terms of RMSE. We suggest using a forecast averaging method and extending our analysis to a wide range of commodity prices.

Project Report from the year 2018 in the subject Computer Science - Technical Computer Science, , course: Computer Science, language: English, abstract: Modeling and Forecasting of the financial market have been an attractive topic to scholars and researchers from various academic fields. The financial market is an abstract concept where financial commodities such as stocks, bonds, and precious metals transactions happen between buyers and sellers. In the present scenario of the financial market world, especially in the stock market, forecasting the trend or the price of stocks using machine learning techniques and artificial neural networks are the most attractive issue to be investigated. As Giles explained, financial forecasting is an instance of signal processing problem which is difficult because of high noise, small sample size, non-stationary, and non-linearity. The noisy characteristics mean the incomplete information gap between past stock trading price and volume with a future price. The stock market is sensitive with the political and macroeconomic environment. However, these two kinds of information are too complex and unstable to gather. The above information that cannot be included in features are considered as noise. The sample size of financial data is determined by real-world transaction records. On one hand, a larger sample size refers a longer period of transaction records; on the other hand, large sample size increases the uncertainty of financial environment during the 2 sample period. In this project, we use stock data instead of daily data in order to reduce the probability of uncertain noise, and relatively increase the sample size within a certain period of time. By non-stationarity, one means that the distribution of stock data is various during time changing. Non-linearity implies that feature correlation of different individual stocks is various. Efficient Market Hypothesis was developed by Burton G. Malkiel in 1991.

The aim of this publication is to identify and apply suitable methods for analysing and predicting the time series of gold prices, together with acquainting the reader with the history and characteristics of the methods and with the time series issues in general. Both statistical and econometric methods, and especially artificial intelligence methods, are used in the case studies. The publication presents both traditional and innovative methods on the theoretical level, always accompanied by a case study, i.e. their specific use in practice. Furthermore, a comprehensive comparative analysis of the individual methods is provided. The book is intended for readers from the ranks of academic staff, students of universities of economics, but also the scientists and practitioners dealing with the time series prediction. From the point of view of practical application, it could provide useful information for speculators and traders on financial markets, especially the commodity markets.

Master's Thesis from the year 2019 in the subject Computer Science - Miscellaneous, University of Gondar (Atse Tewodros Cumpas), course: Information technology, language: English, abstract: Agricultural price predictions are an integral component of trade and policy analysis. As the prices of agricultural commodities directly influence the real income of farmers and it also affects the national foreign currency generate. Sesame is highly produced in some tropical and subtropical rain forest Ethiopia region. The thesis is to build a model that can predict market prices of sesame commodity. Based on the complexity of sesame price prediction; the predicting models used for crop are linear regression, support vector machine and neural network models to predict a future price. A data have been taken from the ECX website (www.ecx.com.et) in the interval of January 2013 to March 2019. The total numbers of records selected to the experiments are 5,327 daily prices are used for proposed models. The experimental result had evaluated by RMSE, MSE and CC metrics. We follow six phase CRISP-DM process model for sesame price prediction. The process phase are, business understanding, data understanding, data preparation, modeling, evaluating and deployment.

This book presents the latest research findings, methods and development techniques related to Ubiquitous and Pervasive Computing (UPC) as well as challenges and solutions from both theoretical and practical perspectives with an emphasis on innovative, mobile and internet services. With the proliferation of wireless technologies and electronic devices, there is a rapidly growing interest in Ubiquitous and Pervasive Computing (UPC). UPC makes it possible to create a human-oriented computing environment where computer chips are embedded in everyday objects and interact with physical world. It also allows users to be online even while moving around, providing them with almost permanent access to their preferred services. Along with a great potential to revolutionize our lives, UPC also poses new research challenges.

It is vitally crucial to establish a method that can accurately forecast prices on the stock exchange market because of the influence the stock market has on the country's ability to raise capital and advance its economic growth. On the stock market, a great number of sensitivity factors are connected to price movement, which is why the progressions associated with such a phenomenon are routinely evaluated. Several neural network models have recently been used to forecast stock prices. In this research, the data related to active companies in the stock market was used to evaluate research questions. Also, the neural network technique was used to look at all data from the market index, fuzzy neural network model, and long short-term memory (LSTM) model from 2020 to 2021. Accordingly, this study aims to forecast the stock price and give a dynamic model with fewer errors using integrated factors, the technical, cardinal, and economic assessment of the market index using the neural network technique. This will be accomplished by utilizing the neural network method. The findings demonstrated that if the combined data of basic analytical factors was used further, we would not only have better training and receive better results, but we would also be able to decrease the prediction error.

This paper proposes a new method for predicting jump arrivals in stock markets with high-frequency limit order book data. We introduce a new model architecture, based on Convolutional Long Short-Term Memory with attention, to apply time series representation learning with memory and to focus the prediction attention on the most important features to improve performance. Using order book data on five liquid U.S. stocks, we provide empirical evidence on the efficacy of the proposed approach. We find that the proposed approach with an attention mechanism outperforms the multi-layer perceptron network as well as the convolutional neural network and Long Short-Term memory model. The use of limit order book data was found to improve the performance of the proposed model in jump prediction, either clearly or marginally, depending on the underlying stock.