The study uses Moving Average, On Balance Volume, and KD (Stochastic Oscillator) to analyze that the technical analysis in which the bull or bear stock markets is efficiency. Also, verifies the changes of market efficiency before and after the financial crisis and whether it can earn excess returns or not by using technical analysis. That is, the returns earned by using technical analysis significantly greater than buy and hold which means the efficiency of technical analysis. Nevertheless, the study also aims to realize that whether the returns of the portfolio of technical indicators better than unit indicator. The companies in our samples are selected by the size of market value top 30 companies in the industries of electronic and finance in order to avoid the effect of market micro structure. Our results are as follows: (1) The returns in bear market are significantly higher than bull market by using MA6-144. (2) The MA6-72 and MA6-144 of financial stock before financial crisis, the returns of technical analysis are significantly better than buy and hold. In the other hand, in the electronic stock, we can use MA6-22-250, KD, and OBV to beat the buy and hold strategy and verify that the market efficiency does not exist. (3) The returns which combine of KD and OBV indicators are significantly higher than KD.

This master thesis aims at creating automatic trading system, which consists of design, implementation, optimization and testing, on U.S stock market. The algorithm is based on trend identification using falling and rising price minimums and maximums over a certain time interval. Based on the identified trend, the algorithm places buy or sell orders on the stock exchange, which parameters are calculated using Keltner Channel and Stochastic Oscillator indicators.

The aim of the diploma thesis is to test the own trading strategies on the exchange market and evaluate their success and applicability in practice. In the introduction of the diploma thesis, there are described basic parameters and basis for testing, such as tested indicators, tested time period and chosen currency pairs. In the next part of the thesis, selected indicators are compiled into three trading strategies, which are then tested . The final part consists of evaluating the results of all three trading strategies .

The thesis analyzes the network traffic on a router with open source firmware. First is chosen a software platform, based on compatibility with available equipment. Then are assessed properties necessary for the development of custom applications. Support for various programming languages provided by the SDK, development environment and the available modules and libraries, for working with network interface. Based on these factors is then chose method to realize the program. He is implemented on the OpenWRT firmware in C / C + + using network library pcap. These funds are used to capture and analyze network traffic. Obtained data are processed using methods of technical analysis, namely on the basis of moving averages, Stochastic oscillator and Bollinger bands. Based on results of these methods are generated and verified estimates of traffic. They are based on linear extrapolation, simplified for fixed intervals. The validity of each method is verified on base of the estimated value. Method is verified if estimated value of the traffic volume is in the Bollinger band, which is given by the standard deviation. Each method is tested several times in real traffic with different input parameters. Then is evaluated the influence of parameters on the error rate of methods. Individual methods are compared and evaluated based on the behavior in different scenarios and based on the average relative error.