Spelling suggestions: "subject:"biostatistik."" "subject:"datastatistik.""
531 |
Change point detection with respect to varianceErdtman, Elias January 2023 (has links)
This thesis examines a simple method for detecting a change with respect to the variance in a sequence of independent normally distributed observations with a constant mean. The method filters out observations with extreme values and divides the sequence into equally large subsequences. For each subsequence, the count of extreme values is translated to a binomial random variable which is tested towards the expected number of extremes. The expected number of extremes comes from prior knowledge of the sequence and a specified probability of how common an extreme value should be. Then specifying the significance level of the goodness-of-fit test yields the number of extreme observations needed to detect a change. The approach is extended to a sequence of independent multivariate normally distributed observations by transforming the sequence to a univariate sequence with the help of the Mahalanobis distance. Thereafter it is possible to apply the same approach as when working with a univariate sequence. Given that a change has occurred, the distribution of the Mahalanobis distance of a multivariate normally distributed random vector with zero mean is shown to approximately follow a gamma distribution. The parameters for the approximated gamma distribution depend only on Σ1−1/2 Σ2Σ1−1/2 with Σ1 and Σ2 being the covariance matrices before and after the change has occurred. In addition to the proposed approach, other statistics such as the largest eigenvalue, the Kullback-Leibler divergence, and the Bhattacharyya distance are considered.
|
532 |
Treatment of Missing Observation in Survey data for Official StatisticsPopoola, Yomi January 2023 (has links)
No description available.
|
533 |
En jämförande simuleringsstudie av ekvivaleringsmetoder / A comparative simulation study of equating methodsLjungstedt, Charlotte, Bodin, Amelia January 2023 (has links)
Högskoleprovet används för antagning till universitet och högskola och kan således vara avgörande för en individs framtid. Därför är det viktigt att den erhållna poängen är baserad på individens kunskapsnivå och inte provets svårighetsnivå. För att säkerställa detta används olika ekvivaleringsmetoder som omvandlar råpoängen från provet till ett ekvivalerat poäng, för att sedan omvandla det till ett normerat poäng. I högskoleprovet är den normerade poängen mellan 0.0 och 2.0. Även om det går att se till att proven från år till år ligger på ungefär samma svårighetsnivå går det inte att kontrollera att provtagargrupperna inte skiljer sig åt åren emellan. Denna studies syfte är därför att ta reda på om det finns någon av tre utvalda ekvivaleringsmetoder, linjär ekvivalering, ekvi-percentil ekvivalering och Item Response Theory ekvivalering (IRT) som presterar bättre än de andra vid två scenarion. Det ena scenariot är när två provtagargrupper har samma generella förmåga medan det andra scenariot ekvivalerar två grupper som skiljer sig åt i förmåga. Detta kontrolleras genom att genomföra studien med simulerat data som liknar högskoleprovet. Syftet med att använda simulerat data är att under kontrollerade förhållanden välja provtagarnas förmåga och provens svårighetsgrad. Jämförelser av metoderna görs genom utvärderingsmåtten bias, medelfel och Root Mean Squared Error (RMSE), samt Difference That Matters (DTM). Resultatet visade att samtliga metoder presterar tillräckligt bra när grupperna har samma generella förmåga. Om de däremot inte har samma generella förmåga bör samtliga eller till och med fler metoder användas innan den ekvivalerade poängen tas fram. / The Swedish Scholastic Aptitude Test (SweSAT) is used for college admissions and can therefore be crucial for an individuals’ future. This is one reason why it is important that the test scores are based on the persons’ ability and is not dependent on variation of difficulty between test versions. To ensure that this is the case different equating methods are used to transform the raw score of the test to an equated score. The SweSAT then standardizes the equated score to a scale between 0.0 and 2.0. Even if it is possible to make sure that the test versions from one year to another are at a similar difficulty level it is not possible to account for the difference between the groups taking the test each time. This study aims to investigate whether any of three chosen equating methods, linear equating, equi-percentile equating and Item Response Theory equating (IRT), perform better than any of the others in two scenarios. The scenarios are when the equated groups generally have the same general ability and when the equated groups differ in ability. This is controlled by performing the study on simulated data made to resemble the SweSAT. The purpose of simulating data is to, in a controlled environment, adjust the test takers abilities and the difficulty of each test version. For comparison between the methods this study uses the Difference That Matters (DTM), Standard Error, bias and Root Mean Squared Error (RMSE). The conclusion drawn is that all methods perform decently when the groups have the same ability, although when the groups have a different ability it is more difficult for all the methods to present a decent equated score. Therefore, several equating methods should be performed before setting the final score.
|
534 |
Active Learning applied on text classificationMattiasson, Malin January 2023 (has links)
No description available.
|
535 |
Exchange Rate Analysis Between the U.S. Dollar and the Japanese YenSakiyama, Yuta January 2023 (has links)
The exchange data between the U.S. Dollar and Japanese Yen are analyzed with three models called the Auto-Regressive Integrated Moving- Average (ARIMA) model, the Generalized Auto-Regressive Conditional Heteroscedastic (GARCH) model, and the Fractional Differencing model. We mainly use log-transformed data with one difference taken because the residuals of the ARIMA model with log-transformed data is closer to normal distribution than other residuals of ARIMA models with squareroot- transformation or without transformation. Furthermore, the Akaike Information Criteria (AIC), the Mean Absolute Error (MAE) and the Root Mean Square Error (RMSE) are used to evaluate performances of each models. The Bayesian Information Criterion (BIC) is also used for GARCH models and we do the Ljung-Box tests for ARIMA models. In conclusion, the fractional differencing model is the best one in the three types of models because its MAE and its RMSE are the smallest values in those of all models we make in this analysis.
|
536 |
Klusteranalys av lejonflockar baserat pågenetiska dataNordqvist, Tobias January 2001 (has links)
No description available.
|
537 |
Korrelationen mellan brott och bostadspriser genom multipel regression / The correlation between crime and housing prices through multiple regressionIbragimov, Hamid, Anwar, Rayan January 2023 (has links)
Syftet med detta arbete är att undersöka ifall brott har en statistiskt signifikantkorrelation med bostadspriser i Stockholm. Detta gjordes genom att tillämpaoch utveckla en multipel regressionsmodell. Datan för arbetet hämtades frånSvensk Mäklarstatistik och Brottsförebyggande Rådet. Arbetet täckte perioden2019-2021 och områdena Kungsholmen, Rinkeby-Kista, Farsta och Skärholmen.Resultatet visar att brott hade en signifikant påverkan på bostadspriset ochpåverkat priset negativt. / The purpose of this paper is to examine if there exists a statistically significantcorrelation between crime and the housing price in Stockholm. This was madeby applying and developing a multiple regression model. The data was gatheredfrom Svensk Mäklarstatistik and Brottsförebyggande Rådet. The analysis coversthe period 2019-2021 in the areas Kungsholmen, Rinkeby-Kista, Farsta andSkärholmen. The results showed a statistically significant correlation betweencrime and housing prices and it showed that crime had a negative effect.
|
538 |
Cut-off and coordinated survey samplingCoder Gylling, Kira January 2023 (has links)
The problem of estimation from a sample is well known and its solutions are the foundation for the sample surveys conducted at e.g. national statistical institutes. This dissertation looks at the estimation problem from the perspective of business survey methodology. The focus is on two types of survey sampling: Cut-off sampling and coordinated sampling. We introduce and evaluate new method for estimation in cut-off sampling as well as a tool for drawing coordinated samples in the R programming language.
|
539 |
Cluster Analysis on Football Teams Performance DataThunberg Kalt, Clarence January 2024 (has links)
No description available.
|
540 |
Sustainable Investments - The impact of the EU green taxonomy / Hållbara Investeringar - Effekten av EU's gröna taxonomiAbrahamsson, Ville, Ekblom, Julia January 2021 (has links)
The increasing environmental issues and the measures taken to tackle them, is a topic of high significance in today's society. In light of this, the EU is underway with developing a taxonomy classifying sustainable economic activities in hopes to raise awareness, increase transparency regarding environmental impact, and motivate investors to invest sustainable. This paper aims to examine if the taxonomy is relevant to its cause, as well as if sustainability factors can be identified with linear regression connected to growth in a company's value, which may motivate sustainable investments. Several interviews were conducted, along with the creation of a mathematical model. The conclusions drawn was that it is not viable to determine a company's growth in value using solely sustainability factors. However, the results were promising regarding the implementation of sustainability factors in more comprehensive models. Furthermore, the impact of the taxonomy was hard to predict at this time, however, the consensus of the majority of the interviews conducted with experts on the subject, is that it has potential to impact sustainable investments in the future. Future research on the taxonomy may yield results of higher interest since more comprehensive data will be available, and the impact of the taxonomy will be more concrete. / De ökande klimatförändringarna och dess åtgärder är ett viktigt ämne i dagens samhälle. I skenet av detta, håller EU på att ta fram en taxonomi som ett verktyg för att klassificera hållbara ekonomiska aktiviteter med hoppet att öka medvetenheten och transparensen kring miljöpåverkan samt motivera investerare att investera hållbart. Syftet med denna studie är att undersöka om taxonomin är relevant för sitt syfte, samt om tillväxt i ett företags marknadsvärde kan relateras till hållbarhetsfaktorer, vilket skulle kunna motivera hållbara investeringar. Flertalet intervjuer genomfördes, tillsammans med skapandet av en matematisk modell. Slutsatserna från modellen var att det inte går att bestämma ett företags värdetillväxt med enbart hållbarhetsfaktorer, däremot såg resultaten från mer omfattande modeller mer lovande ut vad det gäller detta. Vidare var taxonomins inverkan svår att förutspå vid detta tillfälle, däremot var konsensus från majoriteten av de genomförda intervjuerna att den har potential att påverka hållbara investeringar i framtiden. Framtida forskning på taxonomin kan förse mer intressanta resultat eftersom att mer omfattande uppgifter kommer att finnas tillgängligt, och effekterna av taxonomin blir mer konkreta.
|
Page generated in 0.099 seconds