This thesis focuses on the issue of personalization of concrete project. The main goal of the thesis is to broaden this real project, which is dealing with offering study materials to students, with elements of personalization in order to become more useful for its visitors - thanks to displaying of relevant information to each visitor separately. Thesis gives detailed description of the whole implementation process, which was necessary for achievement of this particular state.

Complex diseases such as cancer, cardiovascular diseases and diabetes are often diagnosed too late, which significantly impairs treatment options and, in turn, lowers patient’s survival rate drastically and increases the costs significantly. Moreover, the growth of medical data is faster than the ability of healthcare systems to utilize them. Almost 80% of medical data are unstructured, but they are clinically relevant. On the other hand, technological advancements have made it possible to create different  igital health solutions where healthcare and ICT meet. Also, some individuals have already started to measure their body function parameters, track their health status, research their symptoms and even intervene in treatment options which means a great deal of data is being produced and also indicates that patient-driven health care models are transforming how health care functions. These models include quantified self-tracking, consumer-personalized-medicine and health social networks. This research aims to present an open innovation digital health platform which creates value  y using the overlaps between healthcare, information technology and artificial intelligence. This platform could potentially be utilized for early detection of complex diseases by leveraging Big Data technology which could improve awareness by recognizing pooled symptoms of a specific disease. This would enable individuals to effortlessly and quantitatively track and become aware of changes in their health, and through a dialog with a doctor, achieve diagnosis at a significantly earlier stage. This thesis focuses on a case study of the platform for detecting breast cancer at a  ignificantly earlier stage. A qualitative research method is implemented through reviewing the literature, determining the knowledge gap, evaluating the need, performing market research, developing a conceptual prototype and presenting the open innovation platform. Finally, the value creation, applications and challenges of such platform are investigated, analysed and discussed based on the collected data from interviews and surveys. This study combines an explanatory and an analytical research approach, as it aims not only to describe the case, but also to explain the value creation for different stakeholders in the value chain. The findings indicate that there is an urgent need for early diagnosis of complex diseases such as breast cancer) and also handling direct and indirect consequences of late diagnosis. A significant outcome of this research is the conceptual prototype which was developed based on the general proposed concept through a customer development process. According to the conducted surveys, 95% of the cancer patients and 84% of the healthy individuals are willing to use the proposed platform. The results indicate that it can create significant values for patients, doctors, academic institutions, hospitals and even healthy individuals.

La médecine personnalisée représente une grande attente et un grand espoir dans la lutte contre le cancer. Cette approche vise à adapter les traitements aux caractéristiques personnelles du patient, principalement des biomarqueurs génétiques. Dans notre premier travail, nous avons analysé l'ensemble des médicaments approuvés par la FDA avec un biomarqueur pharmacogénétique dans leur label et montré (1) que l'oncologie représentait un tiers des médicaments avec un biomarqueur dans leur notice et (2) qu'une part importante des médicaments en oncologie mentionnaient le biomarqueur pour requérir un test avant la prescription du médicament contrairement aux autres domaines thérapeutiques. Notre deuxième travail a analysé les essais cliniques soumis à la FDA en vue de la mise sur le marché des thérapies ciblées pour lesquelles il existait une indication restreinte aux patients présentant un certain biomarqueur. Nous concluons que dans deux tiers des cas, l'utilisation du biomarqueur pour sélectionner les patients à traiter était basée sur les résultats des essais cliniques restreints aux patients biomarqueur-positifs et, qu'ainsi, il n'existait aucune donnée clinique permettant de conclure à une différence d'effet traitement selon les valeurs du biomarqueur. Pour notre troisième travail, nous avons réalisé une cartographie de l'ensemble des essais enregistrés sur le registre américain des essais cliniques pour les médicaments anti-cancéreux avec la mention d'un biomarqueur dans leur label. Nous avons mis en évidence des variations très importantes entre les médicaments quant au recours à des essais enrichis et au fait de tester un médicament dans plusieurs indications ou avec plusieurs biomarqueurs prédictifs. Dans notre quatrième travail, nous avons étudié la question du bénéfice apporté par les médicaments anti-cancéreux dans un contexte d'augmentation très importante des prix et grâce à la publication récente de deux échelles par les sociétés européenne et américaine d'oncologie (ESMO et ASCO). Nous avons analysé le bénéfice de tous les médicaments anti-cancéreux approuvés entre 2000 et 2015 pour le traitement d'une tumeur solide. Nous avons montré (1) la faible valeur des récents médicaments anti-cancéreux, (2) l'absence de relation entre le prix et la valeur de ces médicaments et (3) l'absence de différence de bénéfice entre médicaments de médecine personnalisée et médicaments classiques. En conclusion, la présence de biomarqueurs prédictifs dans le label des médicaments---souvent citée comme critère de succès de la médecine personnalisée---est pour l'instant restreinte en grande partie à l'oncologie. Le niveau de preuve pour l'utilité clinique est souvent inconnu car les études sont restreintes à un sous-groupe de patients positifs pour le biomarqueur dès les phases initiales du développement du médicament. Enfin, seul un tiers des médicaments anti-cancéreux approuvés par la FDA entre 2000 et 2015 ont un bénéfice cliniquement pertinent, sans différence de bénéfice clinique entre les médicaments avec et sans biomarqueur et sans relation entre le prix et le bénéfice de ces médicaments. / Personalized medicine represents great expectations and hopes in oncology. This approach aims to adapt treatments to the personal characteristics of the patient, mainly genetic biomarkers. In our first work, we analyzed all the FDA-approved drugs with a pharmacogenetic biomarker in their label and showed (1) that oncology represented one-third of the drugs with a biomarker in their label and (2) a significant portion of oncology drugs mentioned the biomarker to require a biomarker test, contrary to other therapeutic areas. Our second work analyzed the clinical trials submitted to the FDA for the approval of targeted therapies for which there was a indication restricted to biomarker-positive patients. We conclude that in two-thirds of the cases, the use of the biomarker to select the patients to be treated was based on the results of the clinical trials restricted to the biomarker-positive patients. Thus, in these cases, there was no clinical evidence to conclude to a treatment-by-biomarker interaction. For our third work, we mapped all the trials recorded on the US ClinicalTrials.gov registry for anti-cancer drugs with a biomarker labeling. We found very important variations between drugs in the use of enriched trials and in testing of the drug in several indications or with several predictive biomarkers. In our last work, we examined the benefit of anti-cancer drugs in a context of very significant price increases and the recent publication of two scales by the European and American oncology societies (ESMO and ASCO). We analyzed the benefit of all anti-cancer drugs approved between 2000 and 2015 for the treatment of a solid tumor. We have shown (1) the low value of recent anti-cancer drugs, (2) the lack of relationship between the price and the value of these drugs, and (3) the lack of difference of benefice between personalized and “classical” medicines. In conclusion, the presence of predictive biomarkers in the label of drugs --- often cited as a criterion of success of personalized medicine --- is, at least for now, being restricted in large part to oncology. The level of evidence for clinical utility is often unknown because studies are restricted to the subgroup of biomarker-positive patients from the initial stages of the drug development. Finally, only one third of the anti-cancer drugs approved by the FDA between 2000 and 2015 have meaningful clinical benefit and there is no difference in clinical benefit between drugs with and without biomarkers and no relation between the price and the benefit of anti-cancer drugs.

Actualmente, existen varias tiendas por departamentos y tiendas online las cuales ofrecen prendas de vestir de las mejores marcas y tendencias. Sin embargo, es muy común escuchar a las mujeres que no están conformes con las prendas que se venden en dichos lugares, ya sea por el color o modelo de la prenda. Además, este inconveniente les genera largas colas en los vestidores, y tener que esperar mucho tiempo para pagar el producto. Ante dicha problemática, Musas Factory viene a ofrecer a las mujeres un servicio de confección personalizada y envío de prendas mediante un aplicativo móvil. El mercado potencial que se ha estimado es 36,988 personas. Para ello, primeramente se elaboró un Business Model Canvas, en el cual se pudo analizar el giro del negocio desde varios aspectos. Además, se determinó el tamaño de mercado del negocio, lo cual fue un punto importante ya que se pudo corroborar si es que existía demanda del negocio que se propone. En relación a las validaciones se decidió plantear 5 hipótesis para validar los cuadrantes del Business Model Canvas, y mediante el concierge, se determinó la intención de compra. También, se optó por elaborar los planes estratégicos vinculados a las decisiones de producto, precio, plaza y promoción. Para concluir con el trabajo, se efectuó los planes financieros, en el cual se hizo uso de un análisis proyectado para tres años y así como también se calculo los ratios para evaluar la rentabilidad y viabilidad del proyecto. / "Currently, there are several department stores and online stores which offer clothing from the best brands and trends. However, it is very common to hear that women are not satisfied with the clothes sold in these places, either because of the color or model of the garment. In addition, this inconvenience generates long queues in the locker rooms, and having to wait a long time to pay for the product. Faced with this problem, Musas Factory comes to offer women a personalized clothing service and garment delivery using a mobile application. The estimated potential market is 36,988 people. For this, firstly a Business Model Canvas was developed, in which the business could be analyzed from various aspects. In addition, the market size of the business was determined, which was an important point since it could corroborate if there was a demand for the proposed business. Regarding the validations, it was decided to propose 5 hypotheses to validate the quadrants of the Business Model Canvas; and through the concierge, the purchase intention was determined. Also, it was decided to elaborate the strategic plans related to product, price, place and promotion decisions. To conclude the work, the financial plans were made, in which a projected analysis for three years was used, and the ratios were also calculated to evaluate the profitability and viability of the project. / Trabajo de investigación

The ultimate goal of personalized or precision medicine is to identify the best treatment for each patient. An N-of-1 trial is a multiple-period crossover trial performed within a single individual, which focuses on individual outcome instead of population or group mean responses. As in a conventional crossover trial, it is critical to understand carryover effects of the treatment in an N-of-1 trial, especially in situations where there are no washout periods between treatment periods and high volume of measurements are made during the study. Existing statistical methods for analyzing N-of-1 trials include nonparametric tests, mixed effect models and autoregressive models. These methods may fail to simultaneously handle measurements autocorrelation and adjust for potential carryover effects. Distributed lag model is a regression model that uses lagged predictors to model the lag structure of exposure effects. In the dissertation, we first introduce a novel Bayesian distributed lag model that facilitates the estimation of carryover effects for single N-of-1 trial, while accounting for temporal correlations using an autoregressive model. In the second part, we extend the single N-of-1 trial model to multiple N-of-1 trials scenarios. In the third part, we again focus on single N-of-1 trials. But instead of modeling comparison with one treatment and one placebo (or active control), multiple treatments and one placebo (or active control) is considered. In the first part, we propose a Bayesian distributed lag model with autocorrelated errors (BDLM-AR) that integrate prior knowledge on the shape of distributed lag coefficients and explicitly model the magnitude and duration of carryover effect. Theoretically, we show the connection between the proposed prior structure in BDLM-AR and frequentist regularization approaches. Simulation studies were conducted to compare the performance of our proposed BDLM-AR model with other methods and the proposed model is shown to have better performance in estimating total treatment effect, carryover effect and the whole treatment effect coefficient curve under most of the simulation scenarios. Data from two patients in the light therapy study was utilized to illustrate our method. In the second part, we extend the single N-of-1 trial model to multiple N-of-1 trials model and focus on estimating population level treatment effect and carryover effect. A Bayesian hierarchical distributed lag model (BHDLM-AR) is proposed to model the nested structure of multiple N-of-1 trials within the same study. The Bayesian hierarchical structure also improve estimates for individual level parameters by borrowing strength from the N-of-1 trials of others. We show through simulation studies that BHDLM-AR model has best average performance in terms of estimating both population level and individual level parameters. The light therapy study is revisited and we applied the proposed model to all patients’ data. In the third part, we extend BDLM-AR model to multiple treatments and one placebo (or active control) scenario. We designed prior precision matrix on each treatment. We demonstrated the application of the proposed method using a hypertension study, where multiple guideline recommended medications were involved in each single N-of-1 trial.

Background: Companies collect private data about consumers for marketingpurposes. Mobile devices provide marketers with valuable insights intoconsumers' hyper-context information concerning specific consumersituations such as location, time, and environment. Consumers, on the otherhand, may have a different attitude towards how marketers use their privatedata. Purpose: The purpose of this thesis is to analyse and gain insight into theemotional experience of consumers when receiving personalized marketingmessages that are designed using their private data collected from mobiledevices. Method: A qualitative study was implemented and primary data werecollected from semi-structured interviews. The conducted data were analysedto understand the underlying concepts of this thesis research question. Anabductive approach was used and involved back-and-forth engagement withthe empirical findings of the social world for theoretical ideas and with thesecondary sources of literature. Conclusion: Consumers’ can experience both positive and negative feelingssimultaneously. The authors conclude that it was not the personalizedmarketing messages themselves that created negative emotions amongconsumers, it was the knowledge and the feeling of knowing that theadvertisements were created from their 'private data. The positive emotionswere connected to the benefits of receiving advertisements that matched theirinterests.

Un nouveau paradigme tente de s’imposer en oncologie ; identifier les anomalies moléculaires dans la tumeur d’un patient, et proposer une thérapie ciblée, en relation avec ces altérations moléculaires. Nous discutons ici des altérations moléculaires considérées pour une orientation thérapeutique, ainsi que de leurs méthodes d’identification : parmi les altérations recherchées, les anomalies de nombre de copies tiennent une place importante, et nous nous concentrons plus précisément sur leur identification par hybridation génomique comparative (aCGH). Nous montrons, d’abord à partir de lignées cellulaires caractérisées, que l’analyse du nombre de copies par aCGH n’est pas triviale et qu’en particulier le choix de la centralisation peut être déterminant ; différentes stratégies de centralisation peuvent conduire à des profils génomiques différents, certains aboutissant à des interprétations erronées. Nous montrons ensuite, à partir de cohortes de patients, qu’une conséquence majeure est de retenir ou non certaines altérations actionnables dans la prise de décision thérapeutique. Ce travail nous a conduit à développer un workflow complet dédié à l’analyse aCGH, capable de prendre en charge les sources de données les plus courantes. Ce workflow intègre les solutions discutées, assure une entière traçabilité des analyses, et apporte une aide à l’interprétation des profils grâce à des solutions interactives de visualisation. Ce workflow, dénommé rCH, a été implémenté sous forme d’un package R, et déposé sur le site Bioconductor. Les solutions de visualisation interactives sont disponibles en ligne. Le code de l’application est disponible pour une installation sur un serveur institutionnel. / In oncology, a new paradigm tries to impose itself ; analyzing patient’s tumors, and identifying molecular alterations matching with targeted therapies to guide a personalized therapeutic orientation. Here, We discuss the molecular alterations possibly relevant for a therapeutic orientation, as well as the methods used for their identification : among the alterations of interest, copy number variations are widely used, and we more specifically focus on comparative genomic hybridization (aCGH). We show, using well characterized cell lines, that identification of CNV is not trivial. In particular, the choice for centralizing profiles can be critical, and different strategies for adjusting profiles on a theoretical 2n baseline can lead to erroneous interpretations. Next, we show, using tumor samples, that a major consequence is to include, or miss, targetable alterations within the decision procedure. This work lead us to develop a comprehensive workflow, dedicated to aCGH analysis. This workflow supports the major aCGH platforms, ensure a full traceability of the entire process and provides interactive visualization tools to assist the interpretation. This workflow, called rCGH, has been implemented as a R package, and is available on Bioconductor. The interactive visualization tools are available on line, and are ready to be installed on any institutional server.

Background: As retailing moves towards online shopping the number of online purchases have increased substantially over the last year in Sweden (Klarna Bank AB, 2019). The industry of high-end beauty and skincare products has experienced growth in 2019 (PostNord, 2019), which is of interest to investigate. There lies importance for firms in retrieving knowledge of how the target market thinks and reacts. This research allows for a deeper understanding of the motivations of consumer behavior, which will be of high value for retailers and marketers when further operating in 2020 and entering 2021.    Problem discussion: There is a growing body of literature that examines the motivations of consumer behavior, however, identified gaps have yet to accumulate. Even if online personalized advertising (OPA) is increasingly being used by retailers worldwide, its influence on Swedish females remains unexplored. This exploratory study was undertaken in response to the demand in Sweden and it attempts to draw meaningful connections between consumer response to OPA as well as the value motives explaining consumption behaviors.    Purpose: In order to fill gaps in previous literature, this research sought to build a theory, which will make meaningful sense of observations on Swedish females in the age group 18-35-year-old’s purchase behavior. More specifically, the research will create an understanding of how OPA influences Swedish females, and what values motivate the female consumer when completing a purchase of high-end beauty and skincare products, with and without respect to the influence of OPA.   Method: A qualitative approach with semi-structured in-depth interviews with 19 Swedish female participants in the age group 18-35 were conducted. The general analytical procedure for analyzing the collected data was used, and the data was further compared with previous literature.    Results: This research indicates that participants declare empathy and hedonic values to influence the completion of purchase. Within the category of empathy values, security is a key value when shopping online. When exposed to OPA, the values of being well-respected and self-respect were added to the explanation for consumer motives of purchases of high-end beauty and skincare products online.

The control of type-I error is a focal point for clinical trials. On the other hand, it is also critical to be able to detect a truly efficacious treatment in a clinical trial. With recent success in supervised learning (classification and regression problems), artificial intelligence (AI) and machine learning (ML) can play a vital role in identifying efficacious new treatments. However, the high performance of the AI methods, particularly the deep learning neural networks, requires a much larger dataset than those we commonly see in clinical trials. It is desirable to develop a new ML method that performs well with a small sample size (ranges from 20 to 200) and has advantages as compared with the classic statistical models and some of the most relevant ML methods. In this dissertation, we propose a Similarity-Principle-Based Machine Learning (SBML) method based on the similarity principle assuming that identical or similar subjects should behave in a similar manner. SBML method introduces the attribute-scaling factors at the training stage so that the relative importance of different attributes can be objectively determined in the similarity measures. In addition, the gradient method is used in learning / training in order to update the attribute-scaling factors. The method is novel as far as we know. We first evaluate SBML for continuous outcomes, especially when the sample size is small, and investigate the effects of various tuning parameters on the performance of SBML. Simulations show that SBML achieves better predictions in terms of mean squared errors or misclassification error rates for various situations under consideration than conventional statistical methods, such as full linear models, optimal or ridge regressions and mixed effect models, as well as ML methods including kernel and decision tree methods. We also extend and show how SBML can be flexibly applied to binary outcomes. Through numerical and simulation studies, we confirm that SBML performs well compared to classical statistical methods, even when the sample size is small and in the presence of unmeasured predictors and/or noise variables. Although SBML performs well with small sample sizes, it may not be computationally efficient for large sample sizes. Therefore, we propose Recursive SBML (RSBML), which can save computing time, with some tradeoffs for accuracy. In this sense, RSBML can also be viewed as a combination of unsupervised learning (dimension reduction) and supervised learning (prediction). Recursive learning resembles the natural human way of learning. It is an efficient way of learning from complicated large data. Based on the simulation results, RSBML performs much faster than SBML with reasonable accuracy for large sample sizes.

This study explored how women involved in U.S. politics visually framed themselves on their Instagram pages. While recent research in political communications examined the use of Facebook and Twitter, few studies assessed Instagram's role in the game of politics. Guided by political and visual framing theories, a quantitative content analysis of Instagram posts (N = 1,947) from women involved in U.S. politics was conducted. This examination allowed for an exploration of how these public figures framed themselves on Instagram and the extent to which they shared personal content, despite their varied involvement in U.S. politics. Results showed that: both Democrat and Republican women shared political content more often than personal content; Instagram affords a visual-first emphasis for different political issues; and women most often framed themselves as the credible, ideal stateswoman, while still showcasing their personality. Implications for this study affirm Instagram as a legitimate political communications platform, despite its reputation as a food and travel haven.