Skip directly to content

Coming dissertations at Uppsala university

  • Robust machine learning methods Author: Muhammad Osama Link: http://urn.kb.se/resolve?urn=urn:nbn:se:uu:diva-472453 Publication date: 2022-05-12 11:27

    We are surrounded by data in our daily lives. The rent of our houses, the amount of electricity units consumed, the prices of different products at a supermarket, the daily temperature, our medicine prescriptions, our internet search history are all different forms of data. Data can be used in a wide range of applications. For example, one can use data to predict product prices in the future; to predict tomorrow's temperature; to recommend videos; or suggest better prescriptions. However in order to do the above, one is required to learn a model from data. A model is a mathematical description of how the phenomena we are interested in behaves e.g. how does the temperature vary? Is it periodic? What kinds of patterns does it have? Machine learning is about this process of learning models from data by building on disciplines such as statistics and optimization. 

    Learning models comes with many different challenges. Some challenges are related to how flexible the model is, some are related to the size of data, some are related to computational efficiency etc. One of the challenges is that of data outliers. For instance, due to war in a country exports could stop and there could be a sudden spike in prices of different products. This sudden jump in prices is an outlier or corruption to the normal situation and must be accounted for when learning the model. Another challenge could be that data is collected in one situation but the model is to be used in another situation. For example, one might have data on vaccine trials where the participants were mostly old people. But one might want to make a decision on whether to use the vaccine or not for the whole population that contains people of all age groups. So one must also account for this difference when learning models because the conclusion drawn may not be valid for the young people in the population. Yet another challenge  could arise when data is collected from different sources or contexts. For example, a shopkeeper might have data on sales of paracetamol when there was flu and when there was no flu and she might want to decide how much paracetamol to stock for the next month. In this situation, it is difficult to know whether there will be a flu next month or not and so deciding on how much to stock is a challenge. This thesis tries to address these and other similar challenges.

    In paper I, we address the challenge of data corruption i.e., learning models in a robust way when some fraction of the data is corrupted. In paper II, we apply the methodology of paper I to the problem of localization in wireless networks. Paper III addresses the challenge of estimating causal effect between an exposure and an outcome variable from spatially collected data (e.g. whether increasing number of police personnel in an area reduces number of crimes there). Paper IV addresses the challenge of learning improved decision policies e.g. which treatment to assign to which patient given past data on treatment assignments. In paper V, we look at the challenge of learning models when data is acquired from different contexts and the future context is unknown. In paper VI, we address the challenge of predicting count data across space e.g. number of crimes in an area and quantify its uncertainty. In paper VII, we address the challenge of learning models when data points arrive in a streaming fashion i.e., point by point. The proposed method enables online training and also yields some robustness properties.

  • Språklig stil och stajling bland finlandssvenskar i Stockholm : – ett mobilitetsdialektologiskt perspektiv Author: Malin Löfström Link: http://urn.kb.se/resolve?urn=urn:nbn:se:uu:diva-472417 Publication date: 2022-05-12 11:24

    I avhandlingen undersöker jag språklig stil och stajling bland västnyländska finlandssvenskar i Stockholm i ett material som består av intervjuer och smågruppssamtal med elva deltagare. Avhandlingens syfte är att undersöka den språkliga stilen när deltagarna samtalar med en sverigesvensk person. Syftet är också att söka rimliga förklaringar till varför de enskilda individernas språkliga stil och stajling ser ut som de gör i relation till deras identitet och till den sociala betydelse de tillskriver språk och språkförändring.

    Teoretiskt tar undersökningen avstamp i traditionell variations- och sociolingvistik och i modern sociolingvistisk teori. De två paradigmen speglas i analysmetoder och förklaringsmodeller. All språklig analys utförs på sekvenser ur intervjuerna, medan analys av identitetsaspekter och metaspråk utgår från både intervjuer och smågruppssamtal.

    Sammantaget tyder resultaten på att deltagarnas språkliga stil till övervägande del kan kategoriseras som finlandssvensk. De finlandssvenska varianterna förekommer stabilt i deltagargruppen, och analysen av samvariationen mellan varianter som representerar västnyländska, finlandssvenska och sverigesvenska illustrerar att finlandssvenskan kan sägas utgöra en övergångsvarietet eller en mellanvarietet mellan västnyländska och sverigesvenska. Bilden av att finlandssvenskan dominerar den språkliga stilen speglas även i deltagarnas metaspråkliga resonemang, och tolkningen av dessa tyder på att deltagarnas val att använda finlandssvenska i Stockholm i hög grad kan kopplas till identitetsfaktorer.

    Samtidigt visar variationsanalysen att språkförändring i form av stajling med sverigesvenska språkdrag förekommer hos alla deltagare, men i olika omfattning och med olika resurser hos olika individer. Den metaspråkliga analysen stödjer detta resultat. Den uttalade stajlingen går dels ut på att stajla helheten, dels på att använda sverigesvenska ord. Syftet med stajlingen är enligt metakommentarer främst att underlätta kommunikationen. På individnivå är stajlingen dock mer mångfasetterad och de språkliga valen kan kopplas till faktorer i den personliga identiteten och till den sociala betydelsen som deltagarna förknippar med olika språkdrag i Stockholmskontexten.

    En utvärdering av teori och metod indikerar att kombinationen av variationsanalys och tolkande innehållsanalys varit avgörande för att generera resultat som både ger en utförlig beskrivning av den språkliga stilen hos finlandssvenskar i Stockholm och dessutom ger nyanserade förslag på förklaringar till deltagarnas språkliga val i form av stajling.

  • A Sequence of Essays on Sequences of Auctions Author: Daniel Bougt Link: http://urn.kb.se/resolve?urn=urn:nbn:se:uu:diva-472774 Publication date: 2022-05-12 11:07

    Essay I (with Gagan Ghosh and Heng Liu). The existence of declining prices in sequential auctions is a well-documented empirical pattern. Three explanations that can explain the puzzle are bidders being risk averse, loss averse, or ambiguity averse. We use a data set on bids and prices from sequential auctions of train tickets to confirm the existence of declining prices. We further document bidder behavior that is inconsistent with bidders being risk averse or bidders being loss averse.

Pages