Metalearning - Applications to Data Mining

2009

300 citations recorded by the source corpus.

Subfields: Artificial Intelligence

Citations in the corpus

• Learning to Optimize Neural Nets (2017, neutral): “Methods in this category (Brazdil et al., 2008) aim to learn which base-level learner achieves the best performance on a task.”
• Automated Machine Learning: State-of-The-Art and Open Challenges (2019, neutral): “and hence reducing the needed time to learn new tasks [18].”
• Learning to Optimize (2016, neutral): “This line of work, known as “learning to learn” or “meta-learning” [1, 27, 5, 26], considers the problem of devising methods that can take advantage of knowledge learned on other related tasks to train faster, a problem that is today better known as multi-task learning and transfer learning.”
• Meta-Learning: A Survey (2018, neutral): “There exists a rich body of earlier work (Brazdil et al., 2009; Lemke et al., 2015; Giraud-Carrier, 2008; Luo, 2016) on building meta-models for algorithm selection (Bensusan and Giraud-Carrier, 2000; Pfahringer et al.”
• Initializing Bayesian Hyperparameter Optimization via Meta-Learning (2015, positive): “A promising approach to combat this problem is to apply meta-learning (Brazdil et al. 2008) to the hyperparameter search problem.”
• OpenML: networked science in machine learning (2014, neutral): “First, meta-learning projects such as StatLog [27] and MetaL [8], and benchmarking services such as MLcomp(18) run many algorithms on many datasets on their servers.”
• Efficient and Robust Automated Machine Learning (2015, unknown): “The area of meta-learning [15] mimics this strategy by reasoning about the performance of learning algorithms across datasets.”
• Understanding Continual Learning Settings with Data Distribution Drift Analysis (2021, unknown): “…al., 2011; Fei-Fei et al., 2006), transfer learning (Pratt, 1993; Zhao et al., 2017), knowledge distillation (Hinton et al., 2015), meta-learning (Brazdil et al., 2008; Hospedales et al., 2020), learning from sparse supervision and anomaly/concept-drift/out-of-distribution detection (Gama et…”
• A Survey on Ensemble Learning for Data Stream Classification (2017, unknown): “Despite meta-learning being feasible without an ensemble structure [Brazdil et al. 2008; Gama and Kosina 2009], in this work we focus on meta-learning for ensemble classifiers.”
• Towards Automatically-Tuned Neural Networks (2016, unknown): “Firstly, it uses meta-learning (Brazdil et al., 2008) based on experience on previous datasets to start SMAC from good configurations (Feurer et al.”
• Efficient and Robust Automated Machine Learning (2015, unknown): “The area of meta-learning [15] mimics this strategy by reasoning about the performance of learning algorithms across datasets.”
• Metalearning: a survey of trends and technologies (2013, unknown): “Generally, traditional classification algorithms are very successful in metalearning algorithm selection and can include meta-decision trees (Todorovski and Dzeroski 2003), neural networks, Support Vector Machines or any other classification algorithms, with the k-Nearest Neighbours being another po”
• Combining meta-learning and search techniques to select parameters for support vector machines (2012, unknown): “Considering the first issue, a large amount of work has been developed to define suitable meta-features, including general, statistical and information-theoretic measures, landmarkers and model-based measures (see [24] for a review).”
• Meta-learning with neural networks and landmarking for forecasting model selection an empirical evaluation of different feature sets applied to industry data (2016, unknown): “In this context, meta-learning is a promising approach for model selection [8], [9], [10], [11] describing how to deploy knowledge from past tasks in order to select predictive models for new tasks with certain characteristics.”
• Preprocessor Selection for Machine Learning Pipelines (2018, unknown): “Over the past couple of decades the work on automating the machine learning process, commonly referred to as AutoML, has mainly focused on algorithm selection (Brazdil et al., 2009).”
• Learning Gradient Descent: Better Generalization and Longer Horizons (2017, unknown): “However, there is no agreement on the exact definition of meta-learning, and various concepts have been developed by different authors (Thrun & Pratt, 1998; Vilalta & Drissi, 2002; Brazdil et al., 2008).”
• Meta-Learning (2009, unknown): “Typically, the number of recommended algorithms is significantly smaller than the number of all possible (available) algorithms (Brazdil et al., 2009) [3].”
• Machine Learning - The Art and Science of Algorithms that Make Sense of Data (2012, unknown): “Very often diversity can be interpreted as some kind of variance, an idea that already appeared in (Breiman et al., 1984) and was revisited by Langley (1994), Kramer (1996) and Blockeel, De Raedt and Ramon (1998), among others.”
• An instance level analysis of data complexity (2014, unknown): “Data set measures have been used in meta learning (Brazdil et al. 2009) as well as to understand under what circumstances a particular learning algorithm will perform well (Mansilla and Ho 2004).”