Recommendation systems have been widely utilized in e-commerce settings to aid users through their shopping experiences. The principal advantage of these systems is their ability to narrow down the purchase options in addition to marketing items to customers. However, a number of challenges remain, notably those related to obtaining a clearer understanding of users, their profiles, and their preferences in terms of purchased items. Specifically, recommender systems based on collaborative filtering recommend items that have been rated by other users with preferences similar to those of the targeted users. Intuitively, the more information and ratings collected about the user, the more accurate are the recommendations such systems suggest. In a typical recommender systems database, the data are sparse. Sparsity occurs when the number of ratings obtained by the users is much lower than the number required to build a prediction model. This usually occurs because of the users’ reluctance to share their reviews, either due to privacy issues or an unwillingness to make the extra effort. Grey-sheep users pose another challenge. These are users who shared their reviews and ratings yet disagree with the majority in the systems. The current state-of-the-art typically treats these users as outliers and removes them from the system. Our goal is to determine whether keeping these users in the system may benefit learning. Thirdly, cold-start problems refer to the scenario whereby a new item or user enters the system and is another area of active research. In this case, the system will have no information about the new user or item, making it problematic to find a correlation with others in the system. This thesis addresses the three above-mentioned research challenges through the development of machine learning methods for use within the recommendation system setting. First, we focus on the label and data sparsity though the development of the Hybrid Cluster analysis and Classification learning (HCC-Learn) framework, combining supervised and unsupervised learning methods. We show that combining classification algorithms such as k-nearest neighbors and ensembles based on feature subspaces with cluster analysis algorithms such as expectation maximization, hierarchical clustering, canopy, k-means, and cascade k-means methods, generally produces high-quality results when applied to benchmark datasets. That is, cluster analysis clearly benefits the learning process, leading to high predictive accuracies for existing users. Second, to address the cold-start problem, we present the Popular Users Personalized Predictions (PUPP-DA) framework. This framework combines cluster analysis and active learning, or so-called user-in-the-loop, to assign new customers to the most appropriate groups in our framework. Based on our findings from the HCC-Learn framework, we employ the expectation maximization soft clustering technique to create our user segmentations in the PUPP-DA framework, and we further incorporate Convolutional Neural Networks into our design. Our results show the benefits of user segmentation based on soft clustering and the use of active learning to improve predictions for new users. Furthermore, our findings show that focusing on frequent or popular users clearly improves classification accuracy. In addition, we demonstrate that deep learning outperforms machine learning techniques, notably resulting in more accurate predictions for individual users. Thirdly, we address the grey-sheep problem in our Grey-sheep One-class Recommendations (GSOR) framework. The existence of grey-sheep users in the system results in a class imbalance whereby the majority of users will belong to one class and a small portion (grey-sheep users) will fall into the minority class. In this framework, we use one-class classification to provide a class structure for the training examples. As a pre-assessment stage, we assess the characteristics of grey-sheep users and study their impact on model accuracy. Next, as mentioned above, we utilize one-class learning, whereby we focus on the majority class to first learn the decision boundary in order to generate prediction lists for the grey-sheep (minority class). Our results indicate that including grey-sheep users in the training step, as opposed to treating them as outliers and removing them prior to learning, has a positive impact on the general predictive accuracy.

Ces dernières années, le contenu disponible sur le Web a augmenté de manière considérable dans ce qu’on appelle communément le Web social. Pour l’utilisateur moyen, il devient de plus en plus difficile de recevoir du contenu de qualité sans se voir rapidement submergé par le flot incessant de publications. Pour les fournisseurs de service, le passage à l’échelle reste problématique. L’objectif de cette thèse est d’aboutir à une meilleure expérience utilisateur à travers la mise en place de systèmes de filtrage et de recommandation. Le filtrage consiste à offrir la possibilité à un utilisateur de ne recevoir qu’un sous ensemble des publications des comptes auxquels il est abonné. Tandis que la recommandation permet la découverte d’information à travers la suggestion de comptes à suivre sur des sujets donnés. Nous avons élaboré MicroFilter un système de filtrage passant à l’échelle capable de gérer des flux issus du Web ainsi que RecLand, un système de recommandation qui tire parti de la topologie du réseau ainsi que du contenu afin de générer des recommandations pertinentes. / In the last years, the amount of available data on the social Web has exploded. For the average user, it became hard to find quality content without being overwhelmed with publications. For service providers, the scalability of such services became a challenging task. The aim of this thesis is to achieve a better user experience by offering the filtering and recommendation features. Filtering consists to provide for a given user, the ability of receiving only a subset of the publications from the direct network. Where recommendation allows content discovery by suggesting relevant content producers on given topics. We developed MicroFilter, a scalable filtering system able to handle Web-like data flows and RecLand, a recommender system that takes advantage of the network topology as well as the content in order to provide relevant recommendations.

Addressing the needs of developmental math students has been one of the most challenging problems in higher education. Administrators at a private university were concerned about poor academic performance of math-deficient students and sought to identify factors that influenced students' successful progression from developmental to college-level coursework. The purpose of this retrospective prediction study was to determine which of 7 variables (enrollment in a college success course, math placement results, frequency of use of the developmental resource center, source of tuition payment, student's age, gender, and race/ethnicity) would be predictive of success in developmental math as defined by a final course grade of C or higher. Astin's theory of student involvement and Tinto's theory of student retention formed the theoretical framework for this investigation of 557 first-year students who entered the university during Fall 2013 and Fall 2014. Binary logistic regression analysis was performed. Successful completion of the university's college success course as well as enrollment in introductory/intermediate algebra or intermediate algebra were significant predictors of success in remedial math courses. In addition, the lower the level of developmental math a student was placed in and engaged with, the higher the probability of success in the course. These findings were used to create a policy recommendation for a prescriptive means of ensuring students' early enrollment in developmental math courses and engagement with university resources, which may help students overcome barriers to success in developmental math and lead to positive social change for both the students and university through higher retention and graduation rates.

abstract: Playlists have become a significant part of the music listening experience today because of the digital cloud-based services such as Spotify, Pandora, Apple Music. Owing to the meteoric rise in usage of playlists, recommending playlists is crucial to music services today. Although there has been a lot of work done in playlist prediction, the area of playlist representation hasn't received that level of attention. Over the last few years, sequence-to-sequence models, especially in the field of natural language processing have shown the effectiveness of learned embeddings in capturing the semantic characteristics of sequences. Similar concepts can be applied to music to learn fixed length representations for playlists and the learned representations can then be used for downstream tasks such as playlist comparison and recommendation. In this thesis, the problem of learning a fixed-length representation is formulated in an unsupervised manner, using Neural Machine Translation (NMT), where playlists are interpreted as sentences and songs as words. This approach is compared with other encoding architectures and evaluated using the suite of tasks commonly used for evaluating sentence embeddings, along with a few additional tasks pertaining to music. The aim of the evaluation is to study the traits captured by the playlist embeddings such that these can be leveraged for music recommendation purposes. This work lays down the foundation for analyzing music playlists and learning the patterns that exist in the playlists in an end-to-end manner. This thesis finally concludes with a discussion on the future direction for this research and its potential impact in the domain of Music Information Retrieval. / Dissertation/Thesis / Masters Thesis Computer Science 2019

京都大学 / 0048 / 新制・課程博士 / 博士(情報学) / 甲第19851号 / 情博第602号 / 新制||情||105(附属図書館) / 32887 / 京都大学大学院情報学研究科知能情報学専攻 / (主査)教授 美濃 導彦, 教授 石田 亨, 教授 岡部 寿男 / 学位規則第4条第1項該当 / Doctor of Informatics / Kyoto University / DFAM

No description available.

There exist many real-world applications such as recommendation systems, document retrieval, and computational biology where the correct ordering of instances is of equal or greater importance than predicting the exact value of some discrete or continuous outcome. Learning-to-Rank (LTR) refers to a group of algorithms that apply machine learning techniques to tackle these ranking problems. Despite their empirical success, most existing LTR models are not built to be robust to errors in labeling or annotation, distributional data shift, or adversarial data perturbations. To fill this gap, we develop four LTR frameworks that are robust to various types of perturbations. First, Pairwise Elastic Net Regression Ranking (PENRR) is an elastic-net-based regression method for drug sensitivity prediction. PENRR infers robust predictors of drug responses from patient genomic information. The special design of this model (comparing each drug with other drugs in the same cell line and comparing that drug with itself in other cell lines) significantly enhances the accuracy of the drug prediction model under limited data. This approach is also able to solve the problem of fitting on the insensitive drugs that is commonly encountered in regression-based models. Second, Regression-based Ranking by Pairwise Cluster Comparisons (RRPCC) is a ridge-regression-based method for ranking clusters of similar protein complex conformations generated by an underlying docking program (i.e., ClusPro). Rather than using regression to predict scores, which would equally penalize deviations for either low-quality and high-quality clusters, we seek to predict the difference of scores for any pair of clusters corresponding to the same complex. RRPCC combines these pairwise assessments to form a ranked list of clusters, from higher to lower quality. We apply RRPCC to clusters produced by the automated docking server ClusPro and, depending on the training/validation strategy, we show. improvement by 24%–100% in ranking acceptable or better quality clusters first, and by 15%–100% in ranking medium or better quality clusters first. Third, Distributionally Robust Multi-Output Regression Ranking (DRMRR) is a listwise LTR model that induces robustness into LTR problems using the Distributionally Robust Optimization framework. Contrasting to existing methods, the scoring function of DRMRR was designed as a multivariate mapping from a feature vector to a vector of deviation scores, which captures local context information and cross-document interactions. DRMRR employs ranking metrics (i.e., NDCG) in its output. Particularly, we used the notion of position deviation to define a vector of relevance score instead of a scalar one. We then adopted the DRO framework to minimize a worst-case expected multi-output loss function over a probabilistic ambiguity set that is defined by the Wasserstein metric. We also presented an equivalent convex reformulation of the DRO problem, which is shown to be tighter than the ones proposed by the previous studies. Fourth, Inversion Transformer-based Neural Ranking (ITNR) is a Transformer-based model to predict drug responses using RNAseq gene expression profiles, drug descriptors, and drug fingerprints. It utilizes a Context-Aware-Transformer architecture as its scoring function that ensures the modeling of inter-item dependencies. We also introduced a new loss function using the concept of Inversion and approximate permutation matrices. The accuracy and robustness of these LTR models are verified through three medical applications, namely cluster ranking in protein-protein docking, medical document retrieval, and drug response prediction.

Recommendation systems have significantly influenced user experiences across various industries, yet their application in the insurance sector remains relatively unexplored. This thesis focuses on developing a car insurance recommendation system that implements a `consumers like you' feature. The study initially employs a clustering-based recommendation system due to missing labels in an offline environment. However, challenges emerge, such as determining the optimal number of clusters and managing complex data. Additionally, the inability to effectively update based on feedback and lower predictive performance compared to supervised methods necessitated exploring supervised alternatives. In response, this thesis proposes a methodology where the unsupervised approach simulates consumer behavior in an offline environment. Supervised alternatives are pre-trained on the clustering-based system to replicate it and come with the ability to be fine-tuned based on live traffic. Three supervised alternatives — KNN, XGBoost, and a neural network — are developed and compared. Given the supervised recommendation system adaptability based on feedback, supervised methods can provide more accurate, personalized recommendations in the insurance domain. The XGBoost and neural network-based recommendation systems were able to replicate the unsupervised approach, and their expressive power makes them valid candidate models to further evaluate on live traffic. The thesis concludes with the potential to both improve and adapt these recommendation systems to other insurance types, marking a significant step toward more personalized, user-friendly insurance services.

Background: Tuberculosis (TB) is the leading cause of death from a single infectious agent worldwide. The World Health Organization’s (WHO) Global Tuberculosis (GTB) Programme issues evidence-informed guidelines with recommendations on TB. In an effort to improve the accessibility and use of these guidelines, we developed a new digitized WHO eTB catalogue of recommendations. Objective: The objective of this thesis was to explore stakeholder engagement with WHO TB recommendations. We sought to compare the accessibility of the WHO eTB catalogue to the conventional method of accessing WHO TB recommendations, and to explore the ways in which stakeholder feedback could be incorporated into quality improvement frameworks. Methods: We conducted a two-arm superiority randomized controlled trial through a survey among stakeholders who were past or planned future users of TB guidelines, recommendations, or policy advice. Using a 1:1 ratio, we randomly assigned participants to complete an activity using WHO eTB or the conventional website. We compared outcomes of accessibility, understanding, satisfaction and preference between groups. We incorporated qualitative feedback from free-text boxes into a quality improvement framework. Results: From February 26 to March 24 , 2021, we received 188 survey responses, 110 participants were randomized, and 102 were included in the interim analysis. On average, participants rated the WHO eTB catalogue as more accessible across four domains when compared to the WHO TB website. There was no difference in participant understanding of recommendation strength and certainty, but the ability to locate evidence to decision tables favored WHO eTB. We also received 75 qualitative responses, 47 of which yielded five themes: purpose, navigation, presentation, organization, and outreach. Conclusions: The WHO eTB catalogue of recommendations improved the accessibility of WHO TB recommendations and supporting evidence for stakeholders of interest. Our findings support the continued use, promotion, and quality improvement of the WHO eTB catalogue in the future. / Thesis / Master of Public Health (MPH) / Tuberculosis (TB) is a leading cause of death worldwide. The World Health Organization’s (WHO) Global TB (GTB) Programme offers guidelines with recommendations to help decision-makers use evidence on TB prevention, diagnosis, treatment, and care. With the goals of improving the accessibility and use of these recommendations, the WHO and McMaster University have worked together to develop the WHO eTB catalogue of recommendations. This catalogue allows decision-makers to search, filter, and view WHO TB recommendations. This thesis contributed to this work by exploring feedback from decision-makers to identify whether the goals of the WHO eTB catalogue were achieved. The work included creating and leading a randomized controlled trial that compared the WHO eTB catalogue to the earlier way of accessing these recommendations using the WHO publications website. This thesis also explored ways that this feedback could be used to improve the WHO eTB catalogue in the future.

京都大学 / 新制・課程博士 / 博士(経済学) / 甲第24375号 / 経博第662号 / 新制||経||302(附属図書館) / 京都大学大学院経済学研究科経済学専攻 / (主査)教授 西山 慶彦, 教授 江上 雅彦, 講師 柳 貴英 / 学位規則第4条第1項該当 / Doctor of Economics / Kyoto University / DFAM