Stocks belonging to publicly traded companies is a topic which in society is mystified and by some considered to be an unpredictable phenomenon where you either make an economic loss or gain seemingly by chance. Despite this, there are numerous fields of work where the sole purpose is to predict the movement of stocks in order to maximize economic gain. The purpose of this report was to research whether or not these movements actually can be predicted by the usage of regression analysis. A regression model was constructed where the response variable used was the rate of change of a certain stock over 30 days and numerous different qualitative and quantitative parameters were used as regressors. This full model was then evaluated and improved in order to refine its construction and results yielded in order to present the best possible model. When researching and optimizing the model, it was found that several parameters turned out to be statistically significant for the model. The model itself did however come with some uncertainties in the form of a low R-squared value, meaning that despite the significance of said parameters, it contained a high amount of unrepresented variance. / Aktier tillhörande publikt handlade bolag är i samhället ett mystifierat ämne varvid vissa ser det som ett oförutsägbart fenomen, som kan genera antingen vinst eller förlust, till synes av slumpen. Trots detta finns det flertalet områden vars främsta syfte är att förutspå aktiers prisrörelser med ändamålet att maximera ekonomisk vinning. Syftet med denna rapporten var att studera huruvida dessa prisrörelser faktiskt kan förutspås med hjälp av regressionsanalys. En regressionsmodell skapades där rate of change för flertalet aktier under en period på 30 dagar användes som responsvariabel. Flertalet olika kvantitativa och kvalitativa parametrar för respektive aktie användes som regressorer. Den fullständiga modellen som byggde på all data utvärderades för att sedan förbättras, i syfte att förfina dess uppbyggnad och de resultat den genererade, för att skapa en så bra modell som möjligt.  När modellen studerades och optimerades kunde det konstateras att flertalet parametrar var statistiskt signifikanta för modellen. Modellen hade dock osäkerheter i form av bland annat lågt R-kvadratvärde, vilket innebar att trots statistiskt signifikans i flertalet parametrar, kunde modellen inte förklara en stor del av förekommen varians.

The Thesis deals with the current problem related to the connection of dispersive sources to the distribution network. The advantageous legislative support, however, gives a rapid rise to the sources of this kind. Current pricing philosophy favours in particular the growth of photovoltaic power plants. The introduction is aimed at the dispersive sources locating so as to give a comprehensive solution and better understanding of this issue. The equipments of this kind are located in accordance with legislation in force; however, it is necessary to formulate the possibilities of placing them from the view of rational use of the area with the aim of the public weal accordance of both public and private interests. Talking about connection of the new dispersive sources to the distribution network, it is important to mention the legislation, legal directives and regulations delimitating this kind of trade. This theme is the subject of another chapter of the work. Particular attention is paid to the regulation on conditions of connection to the distribution network. When connecting the sources to the distribution network the biggest technical problem from the view of electrical engineering is undesirable influence on the quality parameters (voltage) in this network. This is in particular the voltage change after connection, then, depending on the type of source, the emission of the current harmonic, sharp voltage change, etc. Considering current rise of the dispersive sources connected to the distribution network, more and more influence will appear in quantity parameters. For this reason there will be a lot of search for the methods of preventing from the negative influence on the network. From all of the above said reasons, the work is aimed at the operation analysis of the part of the distribution network, calculation of the network qualitative parameters and searching for the methods of negative influence rectification owing to switching processes of the distribution system. So as to determine the quantitative parameters we used the E-influence programme. Final paragraphs are devoted to the changes to come not only in legislation but, in general, in the approach of the distribution system operators not only to the ways of another connection of such sources but even the regulation of the dispersive sources already connected.