The financial market is a very complex nonlinear series of time. There have been a lot of opinions in the topic of the predictability of it. The need to predict a next day, week, or month has always existed for the final purpose of making money. The most common way of forecasting this time series is with statistic methods and linear regression models. However, the use of artificial intelligence algorithms may have a better outcome, due to the capability of them to handle nonlinear data. The present thesis will be focused on evaluating the use of artificial intelligence algorithms as forecasters for financial markets stock prices. Two algorithms will be used, Feed-Forward Neural networks and Adaptive Neuro-Fuzzy Inference Systems (ANFIS). All forecasts are made with the purpose of a short term trading strategy. Three stocks will be used as an example of the consistency of the method; Google, Apple and the Mexican stock ALFA. These three stocks have different distributed data and different behavior from the neural networks and ANFIS ¡s expected.

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.

This research examines andanalyzes the use of neural networks as a forecasting tool. Specifically a neural network's ability to predict future trends of Stock Market Indices is tested. Accuracy is compared against a traditional forecasting method, multiple linear regression analysis. Finally, the probability of the model's forecast being correct is calculated using conditional probabilities. While only briefly discussing neural network theory, this research determines the feasibility and practicality of usingneural networks as a forecasting tool for the individual investor. This study builds upon the work done byEdward Gately in his book Neural Networks for Financial Forecasting. This research validates the work of Gately and describes the development of a neural network that achieved a 93.3 percent probability of predicting a market rise, and an 88.07 percent probability of predicting a market drop in the S&P500. It was concluded that neural networks do have the capability to forecast financial markets and, if properly trained, the individual investor could benefit from the use of this forecasting tool.

Offers an alternative technique in forecasting to the traditional techniques used in trading and dealing. The book explains the shortcomings of traditional techniques and shows how neural networks overcome many of the disadvantages of these traditional systems.

This research examines and analyzes the use of neural networks as a forecasting tool. Specifically a neural network's ability to predict future trends of Stock Market Indices is tested. Accuracy is compared against a traditional forecasting method, multiple linear regression analysis. Finally, the probability of the model's forecast being correct is calculated using conditional probabilities. While only briefly discussing neural network theory, this research determines the feasibility and practicality of using neural networks as a forecasting tool for the individual investor. This study builds upon the work done by Edward Gately in his book Neural Networks for Financial Forecasting. This research validates the work of Gately and describes the development of a neural network that achieved a 93.3 percent probability of predicting a market rise, and an 88.07 percent probability of predicting a market drop in the S&P500. It was concluded that neural networks do have the capability to forecast financial markets and, if properly trained, the individual investor could benefit from the use of this forecasting tool.

This is abook about the methods developed byour research team, over a period of 10years, for predicting financial market returns. Thework began in late 1991, at a time when one ofus (Jimmy Shadbolt) had just completed a rewrite of the software used at Econostat by the economics team for medium-term trend prediction of economic indica- tors.Looking for anewproject, itwassuggestedthatwelook atnon-linear modelling of financial markets, and that a good place to start might be with neural networks. One small caveat should be added before we start: we use the terms "prediction" and "prediction model" throughout the book, although, with only such a small amount of information being extracted about future performance, can we really claim to be building predictors at all? Some might saythat the future ofmarkets, especially one month ahead, is too dim to perceive. We think we can claim to "predict" for two reasons. Firstlywedoindeedpredictafewper cent offuturevalues ofcertainassets in terms ofpast values ofcertainindicators, asshown by our trackrecord. Secondly, we use standard and in-house prediction methods that are purely quantitative. Weallow no subjective viewto alter what the models tell us. Thus weare doing prediction, even if the problem isvery hard. So while we could throughout the book talk about "getting a better view of the future" or some such euphemism, we would not be correctly describing what it isweare actually doing. Weare indeed getting abetter view of the future, by using prediction methods.