Filters








753 Hits in 2.6 sec

AutoML using Metadata Language Embeddings [article]

Iddo Drori, Lu Liu, Yi Nian, Sharath C. Koorathota, Jie S. Li, Antonio Khalil Moretti, Juliana Freire, Madeleine Udell
2019 arXiv   pre-print
We use language embeddings from modern NLP to improve state-of-the-art AutoML systems by augmenting their recommendations with vector embeddings of datasets and of algorithms.  ...  The resulting (meta-)AutoML framework improves on the performance of existing AutoML frameworks.  ...  This work makes several contributions to the literature by marrying techniques from NLP with AutoML.  ... 
arXiv:1910.03698v1 fatcat:v7myarsbirbetnog5tlc75oowq

Automatic Componentwise Boosting: An Interpretable AutoML System [article]

Stefan Coors and Daniel Schalk and Bernd Bischl and David Rügamer
2021 arXiv   pre-print
In practice, machine learning (ML) workflows require various different steps, from data preprocessing, missing value imputation, model selection, to model tuning as well as model evaluation.  ...  To demonstrate the frameworks efficacy, we compare autocompboost to other existing systems based on the OpenML AutoML-Benchmark.  ...  Our methodological contribution based on a novel stage-and componentwise boosting algorithm is accompanied by an application as well as benchmark experiments to underline the idea and efficacy of our approach  ... 
arXiv:2109.05583v2 fatcat:pp4ykrqusbahhnro32r66f2jei

Evaluation of Representation Models for Text Classification with AutoML Tools [article]

Sebastian Brändle, Marc Hanussek, Matthias Blohm, Maximilien Kintz
2021 arXiv   pre-print
Automated Machine Learning (AutoML) has gained increasing success on tabular data in recent years.  ...  However, processing unstructured data like text is a challenge and not widely supported by open-source AutoML tools.  ...  Methodology of the Benchmark The following chapter describes the underlying methodology and assumptions of the present benchmark.  ... 
arXiv:2106.12798v2 fatcat:trghh26xjzgzlinm3i3x73w2k4

Ensemble Squared: A Meta AutoML System [article]

Jason Yoo, Tony Joseph, Dylan Yung, S. Ali Nasseri, Frank Wood
2021 arXiv   pre-print
There are currently many barriers that prevent non-experts from exploiting machine learning solutions ranging from the lack of intuition on statistical learning techniques to the trickiness of hyperparameter  ...  Empirically, we show that diversity of each AutoML system is sufficient to justify ensembling at the AutoML system level.  ...  Additional support was provided by UBC's Composites Research Network (CRN), Data Science Institute (DSI) and Support for Teams to Advance Interdisciplinary Research (STAIR) Grants.  ... 
arXiv:2012.05390v3 fatcat:3kffo3t4tzf3zm26zbcjq6khte

Landslide Susceptibility Assessment Using an AutoML Framework

Adrián G. Bruzón, Patricia Arrogante-Funes, Fátima Arrogante-Funes, Fidel Martín-González, Carlos J. Novillo, Rubén R. Fernández, René Vázquez-Jiménez, Antonio Alarcón-Paredes, Gustavo A. Alonso-Silverio, Claudia A. Cantu-Ramirez, Rocío N. Ramos-Bernal
2021 International Journal of Environmental Research and Public Health  
This study presents the first attempt to develop a methodology based on an automatic machine learning (AutoML) framework.  ...  This methodology yields better results than other similar methods because using an AutoML framework allows to focus on the treatment of the data, to better understand input variables and to acquire greater  ...  The methodology used is divided into three phases. Firstly, we developed the variables for the study from the information sources to build the raw dataset.  ... 
doi:10.3390/ijerph182010971 pmid:34682717 fatcat:ssgerbok6vao7ghjep7kyzy32q

AutoML @ NeurIPS 2018 challenge: Design and Results [article]

Hugo Jair Escalante, Wei-Wei Tu, Isabelle Guyon, Daniel L. Silver, Evelyne Viegas, Yuqiang Chen, Wenyuan Dai, Qiang Yang
2019 arXiv   pre-print
This data driven competition asked participants to develop computer programs capable of solving supervised learning problems where the i.i.d. assumption did not hold.  ...  In previous competitions we invited participants to build AutoML systems for a wide range of application domains.  ...  In this recently organized challenge, we aimed to explore areas of AutoML that have not been studied so far, and that are present in almost every possible application of AutoML.  ... 
arXiv:1903.05263v2 fatcat:6ugcmzbchzdyjgy6mraf7l5yei

AMC: AutoML for Model Compression and Acceleration on Mobile Devices [article]

Yihui He, Ji Lin, Zhijian Liu, Hanrui Wang, Li-Jia Li, Song Han
2019 arXiv   pre-print
In this paper, we propose AutoML for Model Compression (AMC) which leverage reinforcement learning to provide the model compression policy.  ...  Model compression is a critical technique to efficiently deploy neural network models on mobile devices which have limited computation resources and tight power budgets.  ...  Methodology We present an overview of our AutoML for Model Compression(AMC) engine in Figure 1 . We aim to automatically find the redundancy for each layer, characterized by sparsity.  ... 
arXiv:1802.03494v4 fatcat:o5mbywco6bhgfpdpzlfr3rhsqe

NASirt: AutoML based learning with instance-level complexity information [article]

Habib Asseiss Neto and Ronnie C. O. Alves and Sergio V. A. Campos
2020 arXiv   pre-print
The proposed methodology relies on the Item Response Theory (IRT) for obtaining characteristics from an instance level, such as discrimination and difficulty, and it is able to define a rank of top performing  ...  AutoML is a machine learning field that aims to generate good performing models in an automated way.  ...  classifications from all trained models, and also to the execution of an AutoML tool named Auto-Keras.  ... 
arXiv:2008.11846v2 fatcat:nckpwtw4knbtnglvxocs5niyny

AutoDispNet: Improving Disparity Estimation With AutoML [article]

Tonmoy Saikia, Yassine Marrakchi, Arber Zela, Frank Hutter, Thomas Brox
2019 arXiv   pre-print
Recent AutoML approaches promise to relieve us from this effort. However, they are mainly designed for comparatively small-scale classification tasks.  ...  In this work, we show how to use and extend existing AutoML techniques to efficiently optimize large-scale U-Net-like encoder-decoder architectures.  ...  AutoML [35] in general and automated neural architecture search (NAS [24] ) in particular promise to relieve us from the manual tweaking effort.  ... 
arXiv:1905.07443v2 fatcat:ye7b4vvpynetnjyb46dovl6yqm

Towards Evaluating Exploratory Model Building Process with AutoML Systems [article]

Sungsoo Ray Hong, Sonia Castelo, Vito D'Orazio, Christopher Benthune, Aecio Santos, Scott Langevin, David Jonker, Enrico Bertini, Juliana Freire
2020 arXiv   pre-print
We analyzed the results in two directions: what types of novel insights the AutoML builders learned from end-users, and (2) how the evaluation methodology helped the builders to understand workflows and  ...  To tackle the challenges in evaluation, we propose an evaluation methodology that (1) guides AutoML builders to divide their AutoML system into multiple sub-system components, and (2) helps them reason  ...  Criticism in Methodologies for Evaluating AutoML Designing a reliable evaluation methodology is a notable challenge in AutoML.  ... 
arXiv:2009.00449v1 fatcat:l2giagcq7zhifbwpbwkkgk3gke

Manas: Mining Software Repositories to Assist AutoML [article]

Giang Nguyen, Johir Islam, Rangeet Pan, Hridesh Rajan
2021 arXiv   pre-print
Recent work on AutoML, more precisely neural architecture search (NAS), embodied by tools like Auto-Keras aims to solve this problem by essentially viewing it as a search problem where the starting point  ...  We propose a novel approach for NAS, where instead of starting from a default CNN model, the initial model is selected from a repository of models extracted from GitHub.  ...  Recent work on AutoML, Mining Software Repositories to Assist AutoML.  ... 
arXiv:2112.03395v1 fatcat:h4dbqn7vrjbulltvmad2yfk2ke

Smart Digital Forensic Framework for Crime Analysis and Prediction using AutoML

Sajith A Johnson, S Ananthakumaran
2021 International Journal of Advanced Computer Science and Applications  
We will use the insights and performance metrics derived from our research to motivate forensic intelligence agencies to exploit the features and capabilities provided by AutoML Smart Forensic Framework  ...  applications.  ...  As the autoML computation pipeline also supports TensorFlow which also enables distributed computing applications [9] .  ... 
doi:10.14569/ijacsa.2021.0120349 fatcat:dunsa4borzef3pi4tg25djqjxm

Using a DEA–AutoML Approach to Track SDG Achievements

Bodin Singpai, Desheng Wu
2020 Sustainability  
The back-propagation neural network (BPNN) is used to validate the outputs of the AutoML. As a result, AutoML can outperform BPNN for regression and classification prediction problems.  ...  This work introduces an integrative method that integrates DEA and AutoML to assess and predict performance in SDGs.  ...  Next, we used these predicted POs from the BPNN and AutoML along with the original inputs to measure performance in 2017 using the DEA application.  ... 
doi:10.3390/su122310124 fatcat:ssnapjnopjgb5pw74oqzl5df5e

AutoML Feature Engineering for Student Modeling Yields High Accuracy, but Limited Interpretability

Nigel Bosch
2021 Zenodo  
Thus, we address research questions regarding the accuracy of models built with AutoML features, how AutoML feature types compare to each other and to expert-engineered features, and how interpretable  ...  Finally, we discuss the tradeoffs between effort and interpretability that arise in AutoML-based student modeling.  ...  Methodological work is also needed to improve the interpretability of existing AutoML features or to research AutoML feature engineering methods that incorporate interpretability constraints.  ... 
doi:10.5281/zenodo.5275314 fatcat:j4ldktuubjbu7lzwlnroddsy4a

Improving Neural Networks for Time Series Forecasting using Data Augmentation and AutoML [article]

Indrajeet Y. Javeri, Mohammadhossein Toutiaee, Ismailcem B. Arpinar, Tom W. Miller, John A. Miller
2021 arXiv   pre-print
This paper presents an easy to implement data augmentation method to significantly improve the performance of such networks.  ...  Our method, Augmented-Neural-Network, which involves using forecasts from statistical models, can help unlock the power of neural networks on intermediate length time-series and produces competitive results  ...  AutoML Automated machine learning (AutoML) refers to the process of automating the application of machine learning to realworld problems.  ... 
arXiv:2103.01992v3 fatcat:pefhjtrzz5fhjesfjysqyjrdc4
« Previous Showing results 1 — 15 out of 753 results