Filters








12,444 Hits in 5.6 sec

Meta-Learning Requires Meta-Augmentation [article]

Janarthanan Rajendran, Alex Irpan, Eric Jang
2020 arXiv   pre-print
We describe both of these forms of metalearning overfitting, and demonstrate that they appear experimentally in common meta-learning benchmarks.  ...  This additional level of learning can be powerful, but it also creates another potential source for overfitting, since we can now overfit in either the model or the base learner.  ...  Acknowledgments and Disclosure of Funding We thank Mingzhang Yin and George Tucker for discussion and help in reproducing experimental results for pose regression experiments.  ... 
arXiv:2007.05549v2 fatcat:ag6nwg3chrbbna2oyqumdgpqdy

Survival Prediction of Heart Failure Patients using Stacked Ensemble Machine Learning Algorithm [article]

S.M Mehedi Zaman, Wasay Mahmood Qureshi, Md. Mohsin Sarker Raihan, Ocean Monjur, Abdullah Bin Shams
2021 arXiv   pre-print
Advancement in data mining techniques using machine learning (ML) models is paving promising prediction approaches.  ...  Moreover, we designed and propose a supervised stacked ensemble learning model that can achieve an accuracy, precision, recall and F1 score of 99.98%.  ...  In decision tree-based machine learning techniques, pruning has been shown to be the most effective method of dealing with overfitting [24] .  ... 
arXiv:2108.13367v1 fatcat:psttoev34rcqnk2le5vlon4lvu

Two Challenges of Correct Validation in Pattern Recognition

Thomas Nowotny
2014 Frontiers in Robotics and AI  
The core methods for pattern recognition have been developed by machine learning experts but due to their broad success, an increasing number of non-experts are now employing and refining them.  ...  In this perspective, I will discuss the challenge of correct validation of supervised pattern recognition systems, in particular when employed by nonexperts.  ...  This trend is driven by scientists who are not necessarily experts in machine learning but want to apply machine learning methods in their own application domain.  ... 
doi:10.3389/frobt.2014.00005 fatcat:fpfzr6ruufdmrk7szm5rhdarqe

Unsupervised Neural Machine Translation for Low-Resource Domains via Meta-Learning [article]

Cheonbok Park, Yunwon Tae, Taehee Kim, Soyoung Yang, Mohammad Azam Khan, Eunjeong Park, Jaegul Choo
2021 arXiv   pre-print
To address this issue, this paper presents a novel meta-learning algorithm for unsupervised neural machine translation (UNMT) that trains the model to adapt to another domain by utilizing only a small  ...  Hence, we extend the meta-learning algorithm, which utilizes knowledge learned from high-resource domains, to boost the performance of low-resource UNMT.  ...  Meta Learning Given a small amount of training data, most of machine learning models are prone to overfitting, thus failing to find a generalizable solution.  ... 
arXiv:2010.09046v2 fatcat:2iquah5dufgwfdnzl2kvq7gkl4

L2AE-D: Learning to Aggregate Embeddings for Few-shot Learning with Meta-level Dropout [article]

Heda Song, Mercedes Torres Torres, Ender Özcan, Isaac Triguero
2019 arXiv   pre-print
(b) We also introduce a simple meta-level dropout technique that reduces meta-level overfitting in several few-shot learning approaches.  ...  In our experiments, we find that this simple technique significantly improves the performance of the proposed method as well as various state-of-the-art meta-learning algorithms.  ...  Overfitting can easily occur using conventional machine learning algorithms in such few-shot regime. To avoid this, we need a learning approach with a high generalisation ability.  ... 
arXiv:1904.04339v1 fatcat:36m2yw5gifconji4ad7tyydryu

Current applications of machine learning in the screening and diagnosis of glaucoma: a systematic review and Meta-analysis

Patrick Murtagh
2020 International Journal of Ophthalmology  
A Meta-analysis of diagnostic accuracy in terms of area under the receiver operating curve (AUROC) was performed.  ...  To compare the effectiveness of two well described machine learning modalities, ocular coherence tomography (OCT) and fundal photography, in terms of diagnostic accuracy in the screening and diagnosis  ...  DISCUSSION Meta-Analysis The findings of this Meta-analysis have indicated that there is no statistically significant difference with respect to machine learning between fundal photos and OCT images in  ... 
doi:10.18240/ijo.2020.01.22 pmid:31956584 pmcid:PMC6942952 fatcat:oykjqjtkf5evlc4nwomyfjhltq

Meta-Regularization by Enforcing Mutual-Exclusiveness [article]

Edwin Pan and Pankaj Rajak and Shubham Shrivastava
2021 arXiv   pre-print
It is the second objective where meta-learning models fail for non-mutually exclusive tasks due to task overfitting.  ...  A direct observable consequence of this memorization is that the meta-learning model simply ignores the task-specific training data in favor of directly classifying based on the test-data input.  ...  The primary contribution of our work consists of: 1. Identification and analysis of task-memorization in black-box and optimization based meta learning models. 2.  ... 
arXiv:2101.09819v1 fatcat:mzhw5agqmfa5zfvgfxq64fmuau

Meta-Learning of Evolutionary Strategy for Stock Trading

Erik Sorensen, Ryan Ozzello, Rachael Rogan, Ethan Baker, Nate Parks, Wei Hu
2020 Journal of Data Analysis and Information Processing  
We found that our meta-learning approach to stock trading earns profits similar to a purely evolutionary algorithm.  ...  If achieved, these benefits expand the flexibility of traditional machine learning to areas where there are small windows of time or data available.  ...  Conflicts of Interest The authors declare no conflicts of interest regarding the publication of this paper.  ... 
doi:10.4236/jdaip.2020.82005 fatcat:odgb5p7qbncodk6k2m4l35x2tu

Company bankruptcy prediction framework based on the most influential features using XGBoost and stacking ensemble learning

Much Aziz Muslim, Yosza Dasril
2021 International Journal of Power Electronics and Drive Systems (IJPEDS)  
The prediction analysis process uses the best feature selection and ensemble learning. The best feature selection is selected using feature importance to XGBoost with a weight value filter of 10.  ...  The ensemble learning method used is stacking. Stacking is composed of the base model and meta learner.  ...  Meta learner is a machine learning algorithm that is used to analyze and combine the results of each base model in order to obtain a better prediction rate from the base model.  ... 
doi:10.11591/ijece.v11i6.pp5549-5557 fatcat:wikphdxuwjfkri6ietmrg227tq

Special Issue: Regularization Techniques for Machine Learning and Their Applications

Theodore Kotsilieris, Ioannis Anagnostopoulos, Ioannis E. Livieris
2022 Electronics  
Its primary goal is to make the machine learning algorithm "learn" and not "memorize" by penalizing the algorithm to reduce its generalization error in order to avoid the risk of overfitting.  ...  Regularization is probably the key to address the challenging problem of overfitting, which usually occurs in high-dimensional learning.  ...  Conflicts of Interest: The guest editors declare no conflict of interest.  ... 
doi:10.3390/electronics11040521 doaj:9ffdcf45e0c84dc58f6cdc8ff87c96af fatcat:zj6cfr2jkngelmu5d4arw3u6yi

Accuracy of Computer-Aided Diagnosis of Melanoma

Vincent Dick, Christoph Sinz, Martina Mittlböck, Harald Kittler, Philipp Tschandl
2019 JAMA dermatology  
The recent advances in the field of machine learning have raised expectations that computer-aided diagnosis will become the standard for the diagnosis of melanoma.  ...  Studies were heterogeneous and substantial risk of bias was found in all but 4 of the 70 studies included in the quantitative analysis.  ...  Overfitting is an inherent problem of machine learning resulting in lack of generalizability, especially if the training set and the test set are different from the group of lesions encountered in clinical  ... 
doi:10.1001/jamadermatol.2019.1375 pmid:31215969 pmcid:PMC6584889 fatcat:h66li4nmu5hebmazif3bs3twdq

A Fusion Framework for Forecasting Financial Market Direction Using Enhanced Ensemble Models and Technical Indicators

Dushmanta Kumar Padhi, Neelamadhab Padhy, Akash Kumar Bhoi, Jana Shafi, Muhammad Fazal Ijaz
2021 Mathematics  
The aim of our study was to propose and select a predictive model whose training and testing accuracy difference was minimal in all stocks.  ...  As a result, a proper model selection might allow investors the freedom to invest in a certain stock in order to successfully control risk and create short-term, sustainable profits.  ...  Acknowledgments: Jana Shafi would like to thank the Deanship of Scientific Research, Prince Sattam bin Abdul Aziz University, for supporting this work.  ... 
doi:10.3390/math9212646 fatcat:galxtxhovjfvtcf2b4hkhlslwu

SIAMCAT: user-friendly and versatile machine learning workflows for statistically rigorous microbiome analyses [article]

Jakob Wirbel, Konrad Zych, Morgan Essex, Nicolai Karcher, Ece Kartal, Guillem Salazar, Peer Bork, Shinichi Sunagawa, Georg Zeller
2020 bioRxiv   pre-print
Based on a large meta-analysis of gut microbiome studies, we optimized the choice of ML algorithms and preprocessing routines for default workflow settings.  ...  Here, we present the SIAMCAT R package, a versatile and user-friendly toolbox for comparative metagenome analyses using machine learning (ML), statistical tests, and visualization.  ...  Costea, and Kersten Breuer for helpful discussions and advice on the implementation of SIAMCAT.  ... 
doi:10.1101/2020.02.06.931808 fatcat:jraubeuycrbt5chykvybdlzzuy

Metappearance: Meta-Learning for Visual Appearance Reproduction [article]

Michael Fischer, Tobias Ritschel
2022 arXiv   pre-print
We suggest to combine both techniques end-to-end using meta-learning: We over-fit onto a single problem instance in an inner loop, while also learning how to do so efficiently in an outer-loop that builds  ...  The second approach does not train a model that generalizes across the data, but overfits to a single instance of a problem, e.g., a flash image of a material.  ...  ACKNOWLEDGMENTS This work was supported by Meta Reality Labs, Grant Nr. 5034015.  ... 
arXiv:2204.08993v1 fatcat:smqkgpmtpfcwzmodzbrlx3tz6u

The rationale for ensemble and meta-algorithmic architectures in signal and information processing

Steven J. Simske
2015 APSIPA Transactions on Signal and Information Processing  
Combined, these examples illustrate a new meta-architectural approach to the creation of machine intelligence systems.  ...  Starting with the mature field of ensemble methods and moving to the more-recently introduced field of meta-algorithmics, systems can be designed which are by nature to specifically incorporate new machine-learning  ...  Regardless, this approach often prevents overfitting, and as the AdaBoost [6, 7] algorithm has certainly proven accurate in a number of machine-learning problems.  ... 
doi:10.1017/atsip.2015.10 fatcat:7jijiyw3wnch3csp7mvxybzjxy
« Previous Showing results 1 — 15 out of 12,444 results