The diploma thesis deals with artificial intelligence utilization for predictions on stock markets.The prediction is unconventionally based on Bayes' probabilistic model theorem and on its based Naive Bayes classifier. I the practical part algorithm is designed. The algorithm uses recognized relations between identifiers of technical analyze. Concretely exponential running averages at 20 and 50 days had been used. The program output is a graphic forecast of future stock development which is designed on ground of relations classification between the identifiers

In power electronics, the switched converter plays a fundamental role in the efficient conversion and dynamical control of electrical energy. Due to the switching operation of these systems, overlaid disturbances come into existence in addition to the desired behavior of the variables, causing deviations in the current and voltages. From a control perspective, these disturbances are of no interest since they cannot be compensated. They can even alter the measurements given to the control system, affecting its behavior. Furthermore, during the control design, averaged models are often used, by which the switching operation is somehow disregarded. They consider instead the average behavior of the system variables. Thus, it is essential that the measuring setup provides a measurement of the average value to the control system. To accomplish this goal, there are in practice different approaches. For example, the disturbances originated by the switching operation can be either suppressed using an analog or digital filter, or the sampling of the variables can be carried out in a suitable manner, synchronous to the carrier of the modulation method. Unfortunately, the use of filters adds an extra phase shift or delay to the control loop, reducing its dynamical performance. Moreover, the synchronous sampling method provides a good approximation of the average value only if certain conditions are met, otherwise a distortion due to aliasing takes place. A method is developed in this work to predict, in every switching cycle, the average value of the system variables in a switched power converter. In this context, the work presents an alternative method to carry out the measurement of the average value, avoiding the principal drawbacks of the standard measuring methods. To achieve this, a suitable model of the converter is used, incorporating the modulation method and the type of analog-to-digital converter, either a conventional sample-and-hold or a sigma-delta converter. The measurement given by the analog-to-digital converter is used to predict the time behavior of the system variables during the present switching period and then to evaluate its average value, before the period is completed. The method allows to obtain simultaneously the average value of currents and voltages, to get rid of the delay introduced by filtering, and to avoid the drawback of sampling in the measurement, i.e. aliasing. In this work, an overview of the standard measuring methods for switched power converters is first presented. The problematics that arise from the sampling process are also discussed. Next, the theoretical grounds of the method are developed and the tools needed to implement it are derived. To illustrate its applicability, the method is used first in DC-DC converters, where the case of the buck converter is analyzed in detail. Similarly, the method is applied to a three-phase two-level voltage source converter. In both cases, simulation results and experimental verification are presented for different operational modes. The usage of the method in open and closed loop is discussed, and its effect in the system behavior is shown. The performance of the prediction method is contrasted with other standard measuring methods.

Power electronic devices have become integral components in modern consumer and transportation industries. Predicting the failure or health status of these devices not only ensures operational safety and prevents catastrophic consequences but also leads to reduced downtime and operational costs. However, failure or health status prediction represents a complex problem marked by numerous intrinsic and extrinsic variables, leading to different lifetimes of devices. Additionally, selecting relevant precursor signals that effectively capture the underlying failure mechanisms and overcoming time-series prediction challenges, such as handling dynamic and non-linear behaviors, are crucial for accurate predictions. In the thesis, three models—Kalman filter (KF), Particle filter (PF), and Autoregressive Integrated Moving Average (ARIMA)—are applied, compared, and evaluated for failure or health status prediction of power electronic devices using Power Cycling (PC) test data for power diodes. Among the models, the KF demonstrates the most significant performance while consuming the least amount of time. The PF achieves the second-best performance and the third-best time consumption. Meanwhile, the in-sample ARIMA model delivers the third-best performance and the second-best time consumption. Finally, the out-of-sample ARIMA model ranked the lowest in both performance and time consumption. These results suggest that dynamic models, specifically the KF and PF, exhibit superior generalization capabilities across different devices. This underscores the potential of dynamic models for enhancing predictive accuracy while optimizing computational efficiency in the context of real-time power electronic device health monitoring. / Effektelektronikkomponenter har blivit integrerade delar av moderna konsument- och transportindustrier. Att förutsäga fel eller hälsotillstånd hos dessa enheter säkerställer inte bara operativ säkerhet och förebygger katastrofala konsekvenser utan leder också till minskad driftstopp och lägre driftskostnader. Dock representerar förutsägelse av fel eller hälsotillstånd en komplex uppgift som kännetecknas av många inbyggda och yttre variabler, vilket leder till olika livslängder för enheterna. Dessutom är det avgörande för noggranna förutsägelser att välja relevanta föregångssignaler som effektivt fångar upp de underliggande felmekanismerna och övervinna utmaningar med tidsberoende prediktion, såsom hantering av dynamiska och icke-linjära beteenden. I avhandlingen tillämpas, jämförs och utvärderas tre modeller - Kalman-filter (KF), partikelfilter (PF) och autoregressiv integrerad rörlig medelvärde (ARIMA) - för förutsägelse av fel eller hälsotillstånd hos effektelektronikkomponenter med hjälp av testdata för effektdioder från Power Cycling (PC). Bland modellerna visar KF den mest betydande prestandan samtidigt som den kräver minst tid. PF uppnår den näst bästa prestandan och den tredje bästa tidsåtgången. Samtidigt ger in-sample ARIMA-modellen den tredje bästa prestandan och den näst bästa tidsåtgången. Slutligen rankades out-of-sample ARIMA-modellen lägst både när det gäller prestanda och tidsåtgång. Dessa resultat tyder på att dynamiska modeller, särskilt KF och PF, uppvisar överlägsna generaliseringsförmågor över olika enheter. Detta understryker potentialen hos dynamiska modeller för att förbättra förutsägelseprecisionen samtidigt som de optimerar beräkningskapaciteten i sammanhanget av övervakning av hälsotillståndet för effektelektronikkomponenter i realtid.

The main objective of the thesis was to investigate the feasibility of accurately and precisely fore- casting the number of both national and provincial bene ciaries of social security grants in South Africa, using simple autoregressive integrated moving average (ARIMA) models. The series of the monthly number of bene ciaries of the old age, child support, foster care and disability grants from April 2004 to March 2010 were used to achieve the objectives of the thesis. The conclusions from analysing the series were that: (1) ARIMA models for forecasting are province and grant-type spe- ci c; (2) for some grants, national forecasts obtained by aggregating provincial ARIMA forecasts are more accurate and precise than those obtained by ARIMA modelling national series; and (3) for some grants, forecasts obtained by modelling the latest half of the series were more accurate and precise than those obtained from modelling the full series. / Mathematical Sciences / M.Sc. (Statistics)

Part I - Level crossings and other level functionals.<br />Part II - Some contributions in statistics of extremes and in statistical mechanics.

在選定樣本期間1975-2009年下，動能操作策略在日本股市無法獲得顯著正報酬。在三個子樣本期間：1975年-1989年、1990年-1999年以及2000年-2009年下也獲得相同結論，顯示日本股市不存在動能效應。 動能操作策略中的贏家、輸家排序，與公司的財務特性有關。整體而言，輸家股票在平均成交量、平均市值上皆小於贏家股票。另外，動能操作策略在日本股市的月報酬並沒有明顯季節性變化。 本論文比較文獻上提出的三種不同動能操作策略：歷史報酬率法、52週高點法與移動平均比率法在日本股市的績效表現。三者在日本股市皆無法獲得顯著報酬。最後，進行動能操作策略的形成期間分析。在持有期間第11個月至第18個月內，日本股市出現價格反轉情形。根據形成期間歷史報酬率高低，採用前17個月至前12個月的六個月累積歷史報酬率作為選股依據，採取反向操作策略，發現日本股市存在價格反轉現象。 / Momentum strategies do not yield significant positive returns in the Japanese stock market in the sample period (1975 to 2009). In the three sub-periods, 1975 to 1989, 1990 to 1999 and 2000 to 2009, it demonstrates the same conclusion. Momentum effect does not exist in the Japanese stock market. This study shows that the ranking order of winners and losers is associated with financial characteristics of firm. Overall, average trading volume and average market value of losers stocks are both smaller than those of winners stocks. In addition, the monthly return of momentum strategies has no significant seasonal pattern in the Japanese stock market. In this study, we compare the performance of three different momentum strategies: JT’s individual stock momentum, the 52-week high and the moving average ratio in the Japanese stock market. All of three strategies in the Japanese stock market cannot receive significant profits. Final section tests the periodical analysis of momentum strategies. When extending the holding period, we can find that Japanese stock market experiences price reversal from the 11th to 18th months. According to the historical return in formation period, we choose six-month accumulated historical return (17 to 12 months prior to portfolio formation) as the stock selection principle. Under this contrarian strategy, we find that the Japanese stock market has phenomenon of price reversal.

Statistics / M. Sc. (Statistics)

The main objective of the thesis was to investigate the feasibility of accurately and precisely fore- casting the number of both national and provincial bene ciaries of social security grants in South Africa, using simple autoregressive integrated moving average (ARIMA) models. The series of the monthly number of bene ciaries of the old age, child support, foster care and disability grants from April 2004 to March 2010 were used to achieve the objectives of the thesis. The conclusions from analysing the series were that: (1) ARIMA models for forecasting are province and grant-type spe- ci c; (2) for some grants, national forecasts obtained by aggregating provincial ARIMA forecasts are more accurate and precise than those obtained by ARIMA modelling national series; and (3) for some grants, forecasts obtained by modelling the latest half of the series were more accurate and precise than those obtained from modelling the full series. / Mathematical Sciences / M.Sc. (Statistics)

The idea of time series forecasting techniques is that the past has certain information about future. So, the question of how the information is encoded in the past can be interpreted and later used to extrapolate events of future constitute the crux of time series analysis and forecasting. Several methods such as qualitative techniques (e.g., Delphi method), causal techniques (e.g., least squares regression), quantitative techniques (e.g., smoothing method, time series models) have been developed in the past in which the concept lies in establishing a model either theoretically or mathematically from past observations and estimate future from it. Of all the models, time series methods such as autoregressive moving average (ARMA) process have gained popularity because of their simplicity in implementation and accuracy in obtaining forecasts. But, these models were formulated based on certain properties that a time series is assumed to possess. Classical decomposition techniques were developed to supplement the requirements of time series models. These methods try to define a time series in terms of simple patterns called trend, cyclical and seasonal patterns along with noise. So, the idea of decomposing a time series into component patterns, later modeling each component using forecasting processes and finally combining the component forecasts to obtain actual time series predictions yielded superior performance over standard forecasting techniques. All these methods involve basic principle of moving average computation. But, the developed classical decomposition methods are disadvantageous in terms of containing fixed number of components for any time series, data independent decompositions. During moving average computation, edges of time series might not get modeled properly which affects long range forecasting. So, these issues are to be addressed by more efficient and advanced decomposition techniques such as Wavelets and Empirical Mode Decomposition (EMD). Wavelets and EMD are some of the most innovative concepts considered in time series analysis and are focused on processing nonlinear and nonstationary time series. Hence, this research has been undertaken to ascertain the predictability of nonstationary time series using wavelet and Empirical Mode Decomposition (EMD) based ARMA models. The development of wavelets has been made based on concepts of Fourier analysis and Window Fourier Transform. In accordance with this, initially, the necessity of involving the advent of wavelets has been presented. This is followed by the discussion regarding the advantages that are provided by wavelets. Primarily, the wavelets were defined in the sense of continuous time series. Later, in order to match the real world requirements, wavelets analysis has been defined in discrete scenario which is called as Discrete Wavelet Transform (DWT). The current thesis utilized DWT for performing time series decomposition. The detailed discussion regarding the theory behind time series decomposition is presented in the thesis. This is followed by description regarding mathematical viewpoint of time series decomposition using DWT, which involves decomposition algorithm. EMD also comes under same class as wavelets in the consequence of time series decomposition. EMD is developed out of the fact that most of the time series in nature contain multiple frequencies leading to existence of different scales simultaneously. This method, when compared to standard Fourier analysis and wavelet algorithms, has greater scope of adaptation in processing various nonstationary time series. The method involves decomposing any complicated time series into a very small number of finite empirical modes (IMFs-Intrinsic Mode Functions), where each mode contains information of the original time series. The algorithm of time series decomposition using EMD is presented post conceptual elucidation in the current thesis. Later, the proposed time series forecasting algorithm that couples EMD and ARMA model is presented that even considers the number of time steps ahead of which forecasting needs to be performed. In order to test the methodologies of wavelet and EMD based algorithms for prediction of time series with non stationarity, series of streamflow data from USA and rainfall data from India are used in the study. Four non-stationary streamflow sites (USGS data resources) of monthly total volumes and two non-stationary gridded rainfall sites (IMD) of monthly total rainfall are considered for the study. The predictability by the proposed algorithm is checked in two scenarios, first being six months ahead forecast and the second being twelve months ahead forecast. Normalized Root Mean Square Error (NRMSE) and Nash Sutcliffe Efficiency Index (Ef) are considered to evaluate the performance of the proposed techniques. Based on the performance measures, the results indicate that wavelet based analyses generate good variations in the case of six months ahead forecast maintaining harmony with the observed values at most of the sites. Although the methods are observed to capture the minima of the time series effectively both in the case of six and twelve months ahead predictions, better forecasts are obtained with wavelet based method over EMD based method in the case of twelve months ahead predictions. It is therefore inferred that wavelet based method has better prediction capabilities over EMD based method despite some of the limitations of time series methods and the manner in which decomposition takes place. Finally, the study concludes that the wavelet based time series algorithm could be used to model events such as droughts with reasonable accuracy. Also, some modifications that could be made in the model have been suggested which can extend the scope of applicability to other areas in the field of hydrology.

Many companies consider essential to obtain forecast of time series of uncertain variables that influence their decisions and actions. Marketing includes a number of decisions that depend on a reliable forecast. Forecasts are based directly or indirectly on the information derived from historical data. This data may include different patterns - such as trend, horizontal pattern, and cyclical or seasonal pattern. Most methods are based on the recognition of these patterns, their projection into the future and thus create a forecast. Other approaches such as neural networks are black boxes, which uses learning.