1,720,965 research outputs found
Statistical models and algorithms for data mining and machine learning - Applications to sports analytics and bibliometrics
No full textQuesta tesi di dottorato riassume gli obiettivi di ricerca e le relative risposte
ottenute nell’ambito del programma di dottorato “Modelli e Metodi per
l’Economia e il Management” dell’Università di Brescia, Dipartimento di
Economia e Management, cui l’autore ha partecipato negli anni 2018-2021.
La principale attività di ricerca è relativa all’applicazione delle metodologie
di data science, e in particolare il machine learning, allo sport, focalizzandosi
sul basket, col duplice scopo di fornire allo staff tecnico strumenti per
preparare e analizzare un match, e di affrontare il problema di classificazione
della previsione del risultato di una partita. La seconda attività di ricerca
affrontata è focalizzata sull’utilizzo del linguaggio naturale nel settore bibliometrico,
al fine di offrire strumenti utili alla individuazione e valutazione di
articoli rilevanti per uno specifico contesto scientifico (ad esempio data science
e sport, oppure COVID-19). L’insieme di tecniche utilizzate include, tra
le altre, gli Alberi di Decisione, le Random Forests, il Recursive Partitioning,
il Deep Learning, i Topic Models. Tutte le implementazioni sono scritte nel
linguaggio R, e il codice sorgente è disponibile su richiesta.This PhD dissertation summarizes research questions and related answers
achieved in the context of the PhD program of “Analytics for Economics
and Management” of University of Brescia, Department of Economics and
Management, attended by the author in the years 2018-2021. Main research
activity is related to applying data science methodologies, and machine learning
in particular, to sport, focusing on basketball. Intended goals are both
proposing tools to help coaching staff in matches’ preparation, and facing
the classification problem of outcome prediction of a match. The second
research activity is focused on natural language processing applied to the
bibliometric field, to offer valuable tools in finding and evaluating papers in
a specific research domain (e.g. data science and sport, or COVID-19). The
set of techniques includes, among others, Decision Trees, Random Forests,
Recursive Partitioning, Deep Learning, Topic Models. All implementations
are written in R language, and source code is available on demand
Using Surrogate Models and Variable Importance to better Understand Random Forests Regression Fitting
No full textInterpretability mechanisms helping users in better understanding machine learning models are crucial for Artificial Intelligence acceptance. In this manuscript, our experience in interpretation of random forest regression via surrogate models, i.e. models trying to replicate in an interpretable framework an original fitting difficult to understand, is reported. It is shown how, beyond classical R2 analysis, adequacy of surrogate models can be assessed via variable importance analysis
Integration of model-based recursive partitioning with bias reduction estimation: a case study assessing the impact of Oliver’s four factors on the probability of winning a basketball game
No full textIn this contribution, we investigate the importance of Oliver’s Four Factors, proposed in the literature to identify a basketball team’s strengths and weaknesses in terms of shooting, turnovers, rebounding and free throws, as success drivers of a basketball game. In order to investigate the role of each factor in the success of a team in a match, we applied the MOdel-Based recursive partitioning (MOB) algorithm to real data concerning 19,138 matches of 16 National Basketball Association (NBA) regular seasons (from 2004–2005 to 2019–2020). MOB, instead of fitting one global Generalized Linear Model (GLM) to all observations, partitions the observations according to selected partitioning variables and estimates several ad hoc local GLMs for subgroups of observations. The manuscript’s aim is twofold: (1) in order to deal with (quasi) separation problems leading to convergence problems in the numerical solution of Maximum Likelihood (ML) estimation in MOB, we propose a methodological extension of GLM-based recursive partitioning from standard ML estimation to bias-reduced (BR) estimation; and (2) we apply the BR-based GLM trees to basketball analytics. The results show models very easy to interpret that can provide useful support to coaching staff’s decisions
Going Beyond Counting First Authors in Author Co-citation Analysis
No full textThe present study examines one of the fundamental aspects of author co-citation analysis (ACA) - the way co-citation
counts are defined. Co-citation counting provides the data on which all subsequent statistical analyses and mappings
are based, and we compare ACA results based on two different types of co-citation counting - the traditional type that
only counts the first one among a cited work's authors on the one hand and a non-traditional type that takes into
account the first 5 authors of a cited work on the other hand. Results indicate that the picture produced through this non-traditional author co-citation counting contains more coherent author groups and is therefore considerably clearer. However, this picture represents fewer specialties in the research field being studied than that produced through the traditional first-author co-citation counting when the same number of top-ranked authors is selected and analyzed. Reasons for these effects are discussed
I've seen things you people wouldn't believe: how data science can help coaching staff in basketball
No full textDetecting drivers of basketball successful games: an exploratory study with machine learning algorithms
No full textThis paper aims to detect which are the drivers leading to victory for
basketball matches in NBA, the American National Basketball Association.
First games for regular seasons from 2004-2005 to 2017-2018 have been summarized in terms of box scores and Dean's four factors. Then box scores and four factors have been used as classication independent variables to identify victory drivers, focusing on Golden StateWarriors matches. Both CART and Random Forests machine learning techniques have been applied, and results are compared to assess the more suitable approach
Variations on the Author
No full text“Variations on the Author” discusses two of Eduardo Coutinho’s recent films (Um Dia na Vida, from 2010, and Últimas Conversas, posthumously released in 2015) and their contribution to the general question of documentary authorship. The director’s filmography is characterized by a consistent yet self-effacing form of authorial self-inscription: Coutinho often features as an interviewer that rather than express opinions propels discourses; an interviewer that is good at listening. This mode of self-inscription characterizes him as an author who is not expressive but who is nonetheless markedly present on the screen. In Um Dia na Vida, however, Coutinho is completely absent form the image, while Últimas Conversas, on the contrary, includes a confessional prologue that moves the director from the margins to the center of his films. This article examines the ways in which these works stand out in the filmography of a director who offers new insights into the notion of cinematic authorship
Appropriate Similarity Measures for Author Cocitation Analysis
No full textWe provide a number of new insights into the methodological discussion about author cocitation analysis. We first argue that the use of the Pearson correlation for measuring the similarity between authors’ cocitation profiles is not very satisfactory. We then discuss what kind of similarity measures may be used as an alternative to the Pearson correlation. We consider three similarity measures in particular. One is the well-known cosine. The other two similarity measures have not been used before in the bibliometric literature. Finally, we show by means of an example that our findings have a high practical relevance.information science;Pearson correlation;cosine;similarity measure;author cocitation analysis
Text Mining delle ricerche su COVID-19: Cosa, Chi e Dove
No full textIn questo articolo presentiamo un esempio di come l’applicazione di tecniche statistiche di analisi e visualizzazione dei dati utilizzabili per il cosiddetto “text mining” offra l’opportunità di individuare i temi più rilevanti correlati con l’emergenza causata da COVID-19, le principali fonti informative e le istituzioni dei paesi che si stanno impegnando in queste ricerche
- …
