Generalization Analysis of Fredholm Kernel Regularized Classifiers

2017 ◽  
Vol 29 (7) ◽  
pp. 1879-1901 ◽  
Author(s):  
Tieliang Gong ◽  
Zongben Xu ◽  
Hong Chen
Keyword(s):  
Semisupervised Learning ◽  
Prediction Performance ◽  
Learning Problems ◽  
Kernel Learning ◽  
Classification Algorithms ◽  
Rapid Progress ◽  
Limiting Case ◽  
Fredholm Kernel ◽  
New Framework ◽  
Natural Way

Recently, a new framework, Fredholm learning, was proposed for semisupervised learning problems based on solving a regularized Fredholm integral equation. It allows a natural way to incorporate unlabeled data into learning algorithms to improve their prediction performance. Despite rapid progress on implementable algorithms with theoretical guarantees, the generalization ability of Fredholm kernel learning has not been studied. In this letter, we focus on investigating the generalization performance of a family of classification algorithms, referred to as Fredholm kernel regularized classifiers. We prove that the corresponding learning rate can achieve [Formula: see text] ([Formula: see text] is the number of labeled samples) in a limiting case. In addition, a representer theorem is provided for the proposed regularized scheme, which underlies its applications.

Recent Neuro-Fuzzy Approaches for Feature Selection and Classification

2019 ◽  
pp. 1-19 ◽  
Author(s):  
Heisnam Rohen Singh ◽  
Saroj Kr Biswas ◽  
Monali Bordoloi
Keyword(s):  
Machine Learning ◽  
Feature Selection ◽  
Computation Time ◽  
Prediction Performance ◽  
Learning Problems ◽  
Fuzzy Approach ◽  
Symbolic Forms ◽  
Redundant Data ◽  
Neuro Fuzzy ◽  
And Performance

Classification is the task of assigning objects to one of several predefined categories. However, developing a classification system is mostly hampered by the size of data. With the increase in the dimension of data, the chance of irrelevant, redundant, and noisy features or attributes also increases. Feature selection acts as a catalyst in reducing computation time and dimensionality, enhancing prediction performance or accuracy, and curtailing irrelevant or redundant data. The neuro-fuzzy approach is used for feature selection and classification with better insight by representing knowledge in symbolic forms. The neuro-fuzzy approach combines the merits of neural network and fuzzy logic to solve many complex machine learning problems. The objective of this article is to provide a generic introduction and a recent survey to neuro-fuzzy approaches for feature selection and classification in a wide area of machine learning problems. Some of the existing neuro-fuzzy models are also applied to standard datasets to demonstrate their applicability and performance.

Concept Effect Model

2007 ◽  
pp. 151-170
Author(s):  
Gwo-Jen Hwang
Keyword(s):  
Opportunity To Learn ◽  
Learning Performance ◽  
Learning Problems ◽  
Network Technology ◽  
Rapid Progress ◽  
Adaptive Hypermedia ◽  
Course Content ◽  
Effect Model ◽  
Adaptive Hypermedia Systems ◽  
Relevant Work

With the recent rapid progress of network technology, researchers have attempted to adopt artificial intelligence and use computer networks to develop adaptive hypermedia systems. The idea of adaptive hypermedia is to adapt the course content for a particular learner based on the profile or records of the learner. Meanwhile, researchers have also attempted to develop more effective programs to evaluate the student learning problems, so that the adaptive hypermedia systems can adapt displayed information and dynamically support navigation accordingly. Conventional testing systems simply give students a score, and do not give them the opportunity to learn how to improve their learning performance. Students would benefit more if the test results could be analyzed and hence advice could be provided accordingly. Concept effect model is an effective approach to coping with this problem. In this Chapter, the model and its relevant work are introduced.

scholarly journals A Unifying Framework for Typical Multitask Multiple Kernel Learning Problems

2014 ◽  
Vol 25 (7) ◽  
pp. 1287-1297 ◽  
Author(s):  
Cong Li ◽  
Michael Georgiopoulos ◽  
Georgios C. Anagnostopoulos
Keyword(s):  
Multiple Kernel Learning ◽  
Learning Problems ◽  
Kernel Learning ◽  
Multiple Kernel

scholarly journals Multi-label Problem Transformation Methods: a Case Study

2011 ◽  
Vol 14 (1) ◽  
Author(s):  
Everton Alvares Cherman ◽  
Maria Carolina Monard ◽  
Jean Metz
Keyword(s):  
Classification Problem ◽  
Learning Problems ◽  
Classification Algorithms ◽  
Classification Problems ◽  
Practical Applications ◽  
Problem Transformation ◽  
Textual Data ◽  
Traditional Classification ◽  
Transformation Methods

Traditional classification algorithms consider learning problems that contain only one label, i.e., each example is associated with one single nominal target variable characterizing its property. However, the number of practical applications involving data with multiple target variables has increased. To learn from this sort of data, multi-label classification algorithms should be used. The task of learning from multi-label data can be addressed by methods that transform the multi-label classification problem into several single-label classification problems. In this work, two well known methods based on this approach are used, as well as a third method we propose to overcome some deficiencies of one of them, in a case study using textual data related to medical findings, which were structured using the bag-of-words approach. The experimental study using these three methods shows an improvement on the results obtained by our proposed multi-label classification method.

scholarly journals An Effect of Machine Learning Based Classification Algorithms on Chronic Kidney Disease

2020 ◽  
Vol 9 (3) ◽  
pp. 2249-2256
Keyword(s):  
Machine Learning ◽  
Chronic Kidney Disease ◽  
Kidney Disease ◽  
Prediction Models ◽  
Healthcare Systems ◽  
Medical Data ◽  
Prediction Performance ◽  
Classification Algorithms ◽  
Classification Technique ◽  
Optimal Classifier

In the recent days, the prediction models of chronic kidney disease (CKD) becomes significant in the area of decision making which is helpful in healthcare systems. Because of large amount of medical data, efficient models are required to obtain precise results and data classification algorithms can be employed to detect the presence of CKD. Recently, various machine learning (ML) dependent on data classifier technique is presented for forecasting CKD. Since numerous classification algorithms for CKD prediction exist, there is a need to investigate the prediction performance of these algorithms. This paper propose a comparative analysis of 4 data classifier technique such as deep learning (DL), decision tree (DT), random forest (RF) and random tree (RT). The process of classification technique is analyzed with the help of reputed CKD dataset attained from UCI repository. From the simulation outcomes, it is evident that the DL method achieved optimal classifier action with respect to various namely accuracy, precision and recall.

scholarly journals EVALUATION OF MULTIPLE KERNEL LEARNING ALGORITHMS FOR CROP MAPPING USING SATELLITE IMAGE TIME-SERIES DATA

2017 ◽  
Vol XLII-4/W4 ◽  
pp. 201-207
Author(s):  
S. Niazmardi ◽  
A. Safari ◽  
S. Homayouni
Keyword(s):  
Time Series ◽  
Satellite Image ◽  
Multiple Kernel Learning ◽  
Kernel Learning ◽  
Computational Time ◽  
Classification Algorithms ◽  
Multiple Kernel ◽  
Composite Kernel ◽  
Crop Mapping

Crop mapping through classification of Satellite Image Time-Series (SITS) data can provide very valuable information for several agricultural applications, such as crop monitoring, yield estimation, and crop inventory. However, the SITS data classification is not straightforward. Because different images of a SITS data have different levels of information regarding the classification problems. Moreover, the SITS data is a four-dimensional data that cannot be classified using the conventional classification algorithms. To address these issues in this paper, we presented a classification strategy based on Multiple Kernel Learning (MKL) algorithms for SITS data classification. In this strategy, initially different kernels are constructed from different images of the SITS data and then they are combined into a composite kernel using the MKL algorithms. The composite kernel, once constructed, can be used for the classification of the data using the kernel-based classification algorithms. We compared the computational time and the classification performances of the proposed classification strategy using different MKL algorithms for the purpose of crop mapping. The considered MKL algorithms are: MKL-Sum, SimpleMKL, LPMKL and Group-Lasso MKL algorithms. The experimental tests of the proposed strategy on two SITS data sets, acquired by SPOT satellite sensors, showed that this strategy was able to provide better performances when compared to the standard classification algorithm. The results also showed that the optimization method of the used MKL algorithms affects both the computational time and classification accuracy of this strategy.

Liquid–Liquid Phase Separation in Disease

2019 ◽  
Vol 53 (1) ◽  
pp. 171-194 ◽  
Author(s):  
Simon Alberti ◽  
Dorothee Dormann
Keyword(s):  
Phase Separation ◽  
Living Cells ◽  
Human Diseases ◽  
Therapeutic Interventions ◽  
Rapid Progress ◽  
Disease Mechanisms ◽  
New Concepts ◽  
Condensate Formation ◽  
New Framework ◽  
Liquid Liquid Phase Separation

We have made rapid progress in recent years in identifying the genetic causes of many human diseases. However, despite this recent progress, our mechanistic understanding of these diseases is often incomplete. This is a problem because it limits our ability to develop effective disease treatments. To overcome this limitation, we need new concepts to describe and comprehend the complex mechanisms underlying human diseases. Condensate formation by phase separation emerges as a new principle to explain the organization of living cells. In this review, we present emerging evidence that aberrant forms of condensates are associated with many human diseases, including cancer, neurodegeneration, and infectious diseases. We examine disease mechanisms driven by aberrant condensates, and we point out opportunities for therapeutic interventions. We conclude that phase separation provides a useful new framework to understand and fight some of the most severe human diseases.

scholarly journals Data mining classification algorithms: An overview

2021 ◽  
Vol 8 (2) ◽  
pp. 1-5
Author(s):  
Bardab et al. ◽  
Keyword(s):  
Machine Learning ◽  
Data Mining ◽  
Learning Problems ◽  
Classification Algorithms ◽  
Research Needs ◽  
Logical Relationship

Data mining is also defined as the process of analyzing a quantity of data (usually a large amount) to find a logical relationship that summarizes the data in a new way that is understandable and useful to the owner of the data. This paper examines the various types of classification algorithms in Data Mining, their applications and categorically states the strengths and limitations of each type. The weaknesses found in each algorithm demonstrate how tasks cannot be performed well when only one type of algorithm is applied. For this reason, it is the view of the writer that further research needs to be carried out to explore the potential of combining several of these algorithms to solve machine learning problems.

Sign in / Sign up

Export Citation Format

Share Document