Discovering a taste for the unusual: exceptional models for preference mining

Sá, Cláudio Rebelo de; Duivesteijn, Wouter; Azevedo, Paulo J.; Jorge, Alípio Mário; Soares, Carlos; Knobbe, Arno

doi:10.1007/s10994-018-5743-z

Cited by 13 publications

(6 citation statements)

References 39 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…(1.5, 2.5, −2, 0.5) was converted into (2, 1, 4, 3)). The other three datasets -German2005, German2009 and Sushi were used by de Sa et al [43] and Werbin et al [38]. The Sushi dataset contains the demographics and the preferences of 5000 Japanese over sushi types [44].…”

Section: B Datasetsmentioning

confidence: 99%

BoostLR: A Boosting-Based Learning Ensemble for Label Ranking Tasks

Dery

Shmueli

2020

IEEE Access

View full text Add to dashboard Cite

Label ranking tasks are concerned with the problem of ranking a finite set of labels for each instance according to their relevance. Boosting is a well-known and reliable ensemble technique that was shown to often outperform other learning algorithms. While boosting algorithms were developed for a multitude of machine learning tasks, label ranking tasks were overlooked. Herein, we present a novel boosting algorithm, BoostLR, that was specifically designed for label ranking tasks. Similarly to other boosting algorithms, BoostLR, proceeds in rounds, where in each round, a single weak model is trained over a sampled set of instances. Instances that were identified as harder to predict in the current round, receive a higher (boosted) weight, and therefore also a higher probability to be included in the sample of the forthcoming round. Extensive evaluation of our proposed algorithm on 24 semi-synthetic and real-world label ranking datasets concludes that our algorithm significantly outperforms the current state-of-the-art label ranking methods.

show abstract

Section: B Datasetsmentioning

confidence: 99%

BoostLR: A Boosting-Based Learning Ensemble for Label Ranking Tasks

Dery

Shmueli

2020

IEEE Access

View full text Add to dashboard Cite

show abstract

“…Companies can use diverse consumer-related data (Batistič & der Laken, 2019). However, with the rise of AI and ML algorithms, analyzing data points from multiple data sources to create a holistic view of users is now realistic and attainable (Bąska et al, 2019;Sá et al, 2018).…”

Section: Findings and Discussionmentioning

confidence: 99%

Challenges in Decoding Consumer Behavior with Data Science

Chkoniya¹

2020

Eur. j. econ. bus. stud.

View full text Add to dashboard Cite

Decoding the ever-evolving consumer behavior is one of the biggest challenges faced by marketers around the world. The future of consumer behavior research is put into question by the advances in data science. Today, when consumers are all the time exposed to new technologies, trends such as facial recognition, artificial intelligence, and voice technology did not advance as rapidly as predicted, marketing intelligence gained a significant share of the spotlight. This paper gives an overview of possible ways to anticipate consumer data intelligence development from the perspectives of a robust data set and deep artificial intelligence expertise for better understanding, modeling, and predicting consumer behavior. Showing that marketing cannot happen in isolation in the era of digital overexposure, it requires a deeper understanding of consumer behavior. Data scientists, analysts, and marketers around the world have to work together to increase consumer loyalty, grow revenue, and improve the predictiveness of their models and effectiveness of their marketing spend. Efficiently integrating consumer behavior data into marketing strategies can help companies improve their approach towards attracting and winning the diverse and dynamic consumer segments and retaining them. This synthesis of current research will be helpful to both researchers and practitioners that work on the use of data science to understand and predict consumer behavior, as well as those making long-range planning marketing decisions.

show abstract

“…We also consider two more real datasets, which have also been used in previous works (de Sá et al, 2018;Werbin-Ofir et al, 2019;Dery & Shmueli, 2020). The datasets GermanElec-tions2005 and GermanElections2009 contain socio-economic information from regions of Germany and their respective electoral results.…”

Section: Real Data Applicationsmentioning

confidence: 99%

Comparing Boosting and Bagging for Decision Trees of Rankings

et al. 2021

View full text Add to dashboard Cite

Decision tree learning is among the most popular and most traditional families of machine learning algorithms. While these techniques excel in being quite intuitive and interpretable, they also suffer from instability: small perturbations in the training data may result in big changes in the predictions. The so-called ensemble methods combine the output of multiple trees, which makes the decision more reliable and stable. They have been primarily applied to numeric prediction problems and to classification tasks. In the last years, some attempts to extend the ensemble methods to ordinal data can be found in the literature, but no concrete methodology has been provided for preference data. In this paper, we extend decision trees, and in the following also ensemble methods to ranking data. In particular, we propose a theoretical and computational definition of bagging and boosting, two of the best known ensemble methods. In an experimental study using simulated data and real-world datasets, our results confirm that known results from classification, such as that boosting outperforms bagging, could be successfully carried over to the ranking case.

show abstract

Discovering a taste for the unusual: exceptional models for preference mining

Cited by 13 publications

References 39 publications

BoostLR: A Boosting-Based Learning Ensemble for Label Ranking Tasks

BoostLR: A Boosting-Based Learning Ensemble for Label Ranking Tasks

Challenges in Decoding Consumer Behavior with Data Science

Comparing Boosting and Bagging for Decision Trees of Rankings

Contact Info

Product

Resources

About