The aim of the thesis mainly is to understand an issue of time series analysis. There are many methods in time series analysis, but purpose of this analysis persists the same, which is a construction of sufficient model of time series and his application in forecasting of time series. We have to make a basic identification of time series to establish right process in model constructing. The first and the second chapter is devoted to this basic identification. There are many methods, how we said before, for constructing of concrete model. In this thesis, exactly in the third chapter, we introduce one of the most flexible methodology of model constructing. That is The Box-Jenkins methodology, which was defined in 1976 by these men. In the last chapter we try to put to use insight in the issue of time series analysis for comparison and separation of the space of time series and this comparison use for the right interpretation of the parameters of time series model. The diploma project was supported by project from MSMT of the Czech Republic no. 1M06047 "Centre for Quality and Reliability of Production".

This study is a comparison of forecasting methods for predicting the daily maximum air temperatures in Uppsala using real data from the Swedish Meteorological and Hydrological Institute. The methods for comparison are univariate time series approaches suitable for the data and represent both standard and more recently developed methods. Specifically, three methods are included in the thesis: neural network, ARIMA, and naïve. The dataset is split into a training set and a pseudo out of sample test set. The assessment of which method best forecast the daily temperature in Uppsala is done by comparing the accuracy of the models when doing walk forward validation on the test set. Results show that the neural network is most accurate for the used dataset for both one-step and all multi-step forecasts. Further, the only same-step forecasts from different models that have a statically significant difference are from the neural network and naïve for one- and two-step forecasts, in favor of the neural network.

The objective of the present study is to investigate option pricing and forecasting problems in finance. This is achieved by developing stochastic models in the framework of classical modeling approach. In this study, by utilizing the stock price data, we examine the correctness of the existing Geometric Brownian Motion (GBM) model under standard statistical tests. By recognizing the problems, we attempted to demonstrate the development of modified linear models under different data partitioning processes with or without jumps. Empirical comparisons between the constructed and GBM models are outlined. By analyzing the residual errors, we observed the nonlinearity in the data set. In order to incorporate this nonlinearity, we further employed the classical model building approach to develop nonlinear stochastic models. Based on the nature of the problems and the knowledge of existing nonlinear models, three different nonlinear stochastic models are proposed. Furthermore, under different data partitioning processes with equal and unequal intervals, a few modified nonlinear models are developed. Again, empirical comparisons between the constructed nonlinear stochastic and GBM models in the context of three data sets are outlined. Stochastic dynamic models are also used to predict the future dynamic state of processes. This is achieved by modifying the nonlinear stochastic models from constant to time varying coefficients, and then time series models are constructed. Using these constructed time series models, the prediction and comparison problems with the existing time series models are analyzed in the context of three data sets. The study shows that the nonlinear stochastic model 2 with time varying coefficients is robust with respect different data sets. We derive the option pricing formula in the context of three nonlinear stochastic models with time varying coefficients. The option pricing formula in the frame work of hybrid systems, namely, Hybrid GBM (HGBM) and hybrid nonlinear stochastic models are also initiated. Finally, based on our initial investigation about the significance of presented nonlinear stochastic models in forecasting and option pricing problems, we propose to continue and further explore our study in the context of nonlinear stochastic hybrid modeling approach.

We will compare two different forecasting models with the forecasting model that was used in March 2014 by The Swedish Social Insurance Agency ("Försäkringskassan" in Swedish or "FK") in this degree project. The models are used for forecasting the number of cases. The two models that will be compared with the model used by FK are the Seasonal Exponential Smoothing model (SES) and Auto-Regressive Integrated Moving Average (ARIMA) model. The models will be used to predict case volumes for two types of benefits: General Child Allowance “Barnbidrag” or (BB_ABB), and Pregnancy Benefit “Graviditetspenning” (GP_ANS). The results compare the forecast errors at the short time horizon (22) months and at the long-time horizon (70) months for the different types of models. Forecast error is the difference between the actual and the forecast value of case numbers received every month. The ARIMA model used in this degree project for GP_ANS had forecast errors on short and long horizons that are lower than the forecasting model that was used by FK in March 2014. However, the absolute forecast error is lower in the actual used model than in the ARIMA and SES models for pregnancy benefit cases. The results also show that for BB_ABB the forecast errors were large in all models, but it was the lowest in the actual used model (even the absolute forecast error). This shows that random error due to laws, rules, and community changes is almost impossible to predict. Therefore, it is not feasible to predict the time series with tested models in the long-term. However, that mainly depends on what FK considers as accepted forecast errors and how those forecasts will be used. It is important to mention that the implementation of ARIMA differs across different software. The best model in the used software in this degree project SAS (Statistical Analysis System) is not necessarily the best in other software.

In this paper, we will look at the compatibility of different forecasting methods applied to  time series data in esports, specifically three esports, League of Legends, Counter Strike:Global Offensive and Defence of the Ancients 2. The purpose of the study is to assess whether forecasting the amount of professional esport matches for the first three months of 2021 is possible and if so, how accurately. The forecasting methods used in the report are seasonal ARIMA (SARIMA), autoregressive neural networks (NNAR) and a seasonal naïve model as a benchmark. The results show that, for the chosen methods, all the three datasets were able to fulfill the statistical requirements for producing forecasts as well as outperforming the benchmark model, although with various results. Considering the three games, the one that the study was able to predict with highest accuracy was the CS:GO dataset with a NNAR model where we achieved a mean absolute percentage error of 31%.

År 2020 drabbades Sverige och världen av en pandemin Covid-19. Pandemin har en stor påverkan på flygbranschen enligt tidigare undersökningar. Syftet med studien är att undersöka hur antalet flygpassagerare har påverkats av pandemin samt att jämföra om interventionsmodeller gör mindre prognosfel jämfört med ARIMA-modeller som inte inkluderar en variabel för pandemin. Interventionsanalys av Covid-19 genomförs för att studera effekten av pandemins påverkan på antal flygpassagerare som reser från svenska flygplatser. I mars 2020 gick utrikesdepartementet ut med rekommendation om att undvika onödiga resor för att undvika smittspridning. Interventionsmodeller för inrikes, utrikes och totala antalet flygpassagerare är framtagna. Interventionen betraktas inträffa i mars 2020. Pulsfunktion för maj behöver inkluderas i interventionsmodellen över inrikespassagerare och en pulsfunktion för april behöver modelleras med i interventionsmodellen över utrikespassagerare. För totala antalet flygpassagerare behöver enbart en stegfunktion inkluderas i modellen. Resultaten visar att under covidpandemin har antalet flygpassagerare minskat. Det totala antalet flygpassagerarehar minskat med närmare en miljon passagerare. Utrikespassagerare har minskat med närmare 682000 passagerare och ytterligare cirka 180000 passagerare under lägsta nivåer i april. Inrikespassagerare har minskat med ungefär 370000 passagerare och ytterligare 287000 passagerare i maj. Prognosmodellerna visar delade resultat. För inrikespassagerare blir prognosfelet inte lägre med interventionsmodellerna jämfört med en ARIMA-modell utan interventionseffekt. För utrikespassagerare blir prognosfelet lägre med interventionsmodellerna jämfört med ARIMA-modellen. Över total antalet flygpassagerare gör några av interventionsmodellerna bättre prognoser jämfört med ARIMA-modellen men samtidigt presterade några interventionsmodeller sämre än ARIMA-modellen. / In 2020, Sweden and the world were hit by the Covid-19 pandemic. The pandemic has a major impact on theflight industry according to previous studies. The purpose of this study is to estimate how the number of air passengers has been affected by the pandemic and to estimate models whose purpose is to make short-term forecasts. Intervention analysis is carried out to study the impact of the Covid-19 pandemic on the number of air passengers in Sweden. In March of 2020 the ministry of foreign affairs of Sweden announced a recommendation to avoid unnecessary travels to avoid spreading of the disease. Intervention models for domestic passengers, foreign passengers and the total number of air passengers have been produced. An impulse function for May needed to be included in the intervention model for domestic passengers and an impulse function for April needed to be included in the intervention model for foreign passengers. For the total number of air passengers only a step function for Covid-19 was required. The results show that the Covid-19 pandemic has affected the number of air passenger. The total number of air passengers has decreased by almost one million passengers. Foreign passengers have decreased by almost 682000 passengers and decreased by another 180000 passengers in April 2020. Domestic passengers decreased by approximately 375000 passengers and decreased by another 287000 passengers in May. The forecast models show varying results. For domestic passengers, the forecast errors were not lower for the intervention models compared to the ARIMA model without an intervention effect. For foreign passengers, the forecast errors were lower with the intervention models compared to the ARIMA model. For the total number of passengers, some of the intervention models made better forecasts compared to the ARIMA model, but at the same time some of the intervention models performed worse than the ARIMA model.

Tourism is one of the fastest growing industries in the world, employing approximately 220 million people and generating over 9.4% of the world's GDP. The growing contribution of tourism is accompanied by an increased interest in understanding the major factors which influence visitation levels to those countries. Therefore, finding the right variables to understand and estimate tourism demand becomes very important and challenging in policy formulations. The purpose of this study is to introduce Economic Sentiment Indicator (ESI) to the field of tourism demand studies. Using ESI in demand analysis, this study will assist in the ability to tap into individuals' hopes and/or worries for the present and future. The study developed a demand model in which the number of tourist arrivals to Turkey from select EU countries is used as the dependent variable. ESI along with more traditional variables such as Interest Rate, Relative Price, and Relative Exchange Rate were brought into the model as the independent demand determinants. The study utilized such econometric models as ARIMA for seasonality adjustment and ARDL Bound test approach to cointegration for the long and short-run elasticities. ESI was statistically significant in 8 countries out of 13, three of those countries had a negative coefficient and five had a positive sign as proposed by the study. The study posits that ESI is a good indicator to gauge and monitor tourism demand and adding the visitors' state of mind into the demand equation could reduce errors and increase variance in arrivals. Policy makers should monitor ESI as it fluctuates over time. Since we do not have direct influence on travelers' demand for tourism, it is imperative that we use indirect approaches such as price adjustment and creating new packages or promotional expenditures in order to influence or induce demand. Using this information generated from the study, government officials and tourism suppliers could adjust their promotional activities and expenditures in origin countries accordingly. / Master of Science

The purpose of this study is to add an understanding on the relationship between characteristics of prison population and environmental characteristics of prisons with prison misconduct in the context of Japan, where little empirical research on prison misconduct has been conducted. To aid in the analysis, three theories (the deprivation, importation, and administrative control models) that had been developed in Western countries will be utilized. This study will test thirteen predictors derived from these theories. Using time-series data obtained from annual official reports of Japanese prisons between 1972 and 2019, the relationships between characteristics of prison population (gender, age, violent conviction, criminality, and health problems) and environmental characteristic of prisons or environment-driven characteristics of prisoners (occupancy rate, sentence length, foreign prisoners, drug conviction, and staff-to-inmate ratio) on prison misconduct (total, violent, and non-violent misconduct and refusal to work) were examined through descriptive analysis, graphical portrayal, bivariate correlations, and multivariate analysis using ARIMA (autoregressive integrated moving average) analysis. The results showed that all models can predict prison misconduct partially. Staff-to-inmate ratio was the most consistently significant predictor in this study. Occupancy rate and old age were also significantly related to several types of prison misconduct. Limitations and policy implications are discussed considering these results.

Currently, manufacturing companies/organizations are exceedingly focused on the reverse logistics since it has its own share in the overall profitability and development of the organization. Proper management of the product returns are considered inevitable factor for success by many companies. Warehouses are important part of the reverse logistic chain where major part of the logistic management often require. In order to manage the product returns in an efficient manner in a warehouse, it is very important to have proper planning and proficient method to deal with any uncertain situations. Along with this updating technology, better staff allocation, proper communication etc. are considered as very important for the better function of the product returns management in an organization.  The study was conducted in returns management section of a warehouse facility. The aim of the thesis is to tackle the uncertainty with the help of an efficient forecasting method to predict rate of product returns and further to understand the importance of forecasting in upbringing the performance of the warehouse. The first phase of the study also investigates through the current trend of the rate of reverse flow and proposal of the best suited method for forecasting of the future state. The second aim of the thesis is to improve the current method utilized for managing the product returns in the warehouse and improve the overall cycle time of the system under study. Second phase of the research also focuses towards lean warehousing by eliminating the warehouse wastes in the return management section. Finally, the results obtained in the study is linked with building and improving the key performance indicators (KPI’s) in the return management section of the case company.

The objective of this paper is to retrospectively evaluate forecast models for polling data, to be used prospectively for the Swedish general election in 2022. One of the simplest ways of forecasting an election result is through opinion polls, and using the latest observation as the forecast. This paper considers five different forecasting models on polling data which are evaluated based on different error measures and the results are compared to previous research done on the same topic. The data in this paper consists of time series data of party-preference polls from Statistics Sweden. When forecasting polling data, the naive forecasting model was the most accurate, but forecasting the election in 2018 resulted in the forecast combinations model being the most accurate. Finally, the models are used to make forecasts on the Swedish general election taking place in September of 2022.