Knowledge Combination vs. Meta-Learning

2011 ◽  
pp. 2325-2331
Author(s):  
Ivan Bruha
Keyword(s):  
Knowledge Integration ◽  
Learning Strategy ◽  
Combination Strategy ◽  
General Belief ◽  
Useful Knowledge ◽  
Large Databases ◽  
Meta Learning ◽  
Very Large Databases ◽  
Intelligent Information ◽  
High Level

Research in intelligent information systems investigates the possibilities of enhancing their over-all performance, particularly their prediction accuracy and time complexity. One such discipline, data mining (DM), processes usually very large databases in a profound and robust way (Fayyad et al., 1996). DM points to the overall process of determining a useful knowledge from databases, that is, extracting high-level knowledge from low-level data in the context of large databases. This article discusses two newer directions in this field, namely knowledge combination and meta-learning (Vilalta & Drissi, 2002). There exist approaches to combine various paradigms into one robust (hybrid, multistrategy) system which utilizes the advantages of each subsystem and tries to eliminate their drawbacks. There is a general belief that integrating results obtained from multiple lower-level decision-making systems, each usually (but not required) based on a different paradigm, produce better performance. Such multi-level knowledgebased systems are usually referred to as knowledge integration systems. One subset of these systems is called knowledge combination (Fan et al., 1996). We focus on a common topology of the knowledge combination strategy with base learners and base classifiers (Bruha, 2004). Meta-learning investigates how learning systems may improve their performance through experience in order to become flexible. Its goal is to search dynamically for the best learning strategy. We define the fundamental characteristics of the meta-learning such as bias, and hypothesis space. Section 2 surveys the various directions in algorithms and topologies utilized in knowledge combination and meta-learning. Section 3 represents the main focus of this article: description of knowledge combination techniques, meta-learning, and a particular application including the corresponding flow charts. The last section presents the future trends in these topics.

Interactive Visual Analytics of Databases and Frequent Sets

2013 ◽  
Vol 3 (4) ◽  
pp. 120-140 ◽  
Author(s):  
Carson K.S. Leung ◽  
Christopher L. Carmichael ◽  
Patrick Johnstone ◽  
David Sonny Hung-Cheung Yuen
Keyword(s):  
Visual Analytics ◽  
Journal Article ◽  
Real Life ◽  
Visual Representations ◽  
Analytic Solutions ◽  
Useful Knowledge ◽  
Frequent Sets ◽  
Large Databases ◽  
Advanced Analysis

In numerous real-life applications, large databases can be easily generated. Implicitly embedded in these databases is previously unknown and potentially useful knowledge such as frequently occurring sets of items, merchandise, or events. Different algorithms have been proposed for managing and retrieving useful information from these databases. Various algorithms have also been proposed for mining these databases to find frequent sets, which are usually presented in a lengthy textual list. As “a picture is worth a thousand words”, the use of visual representations can enhance user understanding of the inherent relationships among the mined frequent sets. Many of the existing visualizers were not designed to visualize these mined frequent sets. In this journal article, an interactive visual analytic system is proposed for providing visual analytic solutions to the frequent set mining problem. The system enables the management, visualization, and advanced analysis of the original transaction databases as well as the frequent sets mined from these databases.

scholarly journals A transferable active-learning strategy for reactive molecular force fields

2021 ◽  
Author(s):  
Tom Young ◽  
Tristan Johnston-Wood ◽  
Volker L. Deringer ◽  
Fernanda Duarte
Keyword(s):  
Machine Learning ◽  
Active Learning ◽  
Learning Strategy ◽  
Force Fields ◽  
Molecular Simulations ◽  
Quantum Mechanical ◽  
Interatomic Potentials ◽  
Molecular Force ◽  
High Level ◽  
Active Learning Strategy

Predictive molecular simulations require fast, accurate and reactive interatomic potentials. Machine learning offers a promising approach to construct such potentials by fitting energies and forces to high-level quantum-mechanical data, but...

scholarly journals Knowledge Integration Networks for Action Recognition

2020 ◽  
Vol 34 (07) ◽  
pp. 12862-12869
Author(s):  
Shiwen Zhang ◽  
Sheng Guo ◽  
Limin Wang ◽  
Weilin Huang ◽  
Matthew Scott
Keyword(s):  
Action Recognition ◽  
Large Scale ◽  
Knowledge Integration ◽  
Scene Recognition ◽  
Teacher Networks ◽  
Medium Level ◽  
Meaningful Context ◽  
Context Knowledge ◽  
High Level ◽  
Context Features

In this work, we propose Knowledge Integration Networks (referred as KINet) for video action recognition. KINet is capable of aggregating meaningful context features which are of great importance to identifying an action, such as human information and scene context. We design a three-branch architecture consisting of a main branch for action recognition, and two auxiliary branches for human parsing and scene recognition which allow the model to encode the knowledge of human and scene for action recognition. We explore two pre-trained models as teacher networks to distill the knowledge of human and scene for training the auxiliary tasks of KINet. Furthermore, we propose a two-level knowledge encoding mechanism which contains a Cross Branch Integration (CBI) module for encoding the auxiliary knowledge into medium-level convolutional features, and an Action Knowledge Graph (AKG) for effectively fusing high-level context information. This results in an end-to-end trainable framework where the three tasks can be trained collaboratively, allowing the model to compute strong context knowledge efficiently. The proposed KINet achieves the state-of-the-art performance on a large-scale action recognition benchmark Kinetics-400, with a top-1 accuracy of 77.8%. We further demonstrate that our KINet has strong capability by transferring the Kinetics-trained model to UCF-101, where it obtains 97.8% top-1 accuracy.

scholarly journals Sampling Methods in Approximate Query Answering Systems

2011 ◽  
pp. 990-994 ◽  
Author(s):  
Gautam Das
Keyword(s):  
Data Analysis ◽  
Large Data ◽  
Massive Datasets ◽  
Data Repositories ◽  
Large Databases ◽  
Approximate Query Answering ◽  
Very Large Databases ◽  
Approximate Query ◽  
And Storage ◽  
Collection And Management

In recent years, advances in data collection and management technologies have led to a proliferation of very large databases. These large data repositories typically are created in the hope that, through analysis such as data mining and decision support, they will yield new insights into the data and the real-world processes that created them. In practice, however, while the collection and storage of massive datasets has become relatively straightforward, effective data analysis has proven more difficult to achieve. One reason that data analysis successes have proven elusive is that most analysis queries, by their nature, require aggregation or summarization of large portions of the data being analyzed. For multi-gigabyte data repositories, this means that processing even a single analysis query involves accessing enormous amounts of data, leading to prohibitively expensive running times. This severely limits the feasibility of many types of analysis applications, especially those that depend on timeliness or interactivity.

scholarly journals Towards Religious Understanding: An Assessment of the Ambivalence Roles of Religious Leaders in Christian-Muslim Relations in Nigeria

2021 ◽  
pp. 9-20
Author(s):  
Rapheal Joseph Ojo
Keyword(s):  
Religious Communities ◽  
Ethnographic Research ◽  
Religious Leaders ◽  
General Belief ◽  
Religious Conflicts ◽  
The Government ◽  
High Level ◽  
The Relationship ◽  
Religious Understanding ◽  
Shaky Ground

The world today is becoming more violent than ever before. Sometimes, the violence can be political, ethnic, economic and or religious. In most cases, distinguishing the main cause of such violence from other causes might be difficult. The factors could be a combination of two issues viz: ethnoreligious conflicts or politico-religious conflicts. The religious experience in Nigeria today, as a multi-religious society so far has proven contrary to the general belief and the widespread expectation of people about religion as an institution that promotes social integration. Christian-Muslim relations in Nigeria today (though being the dominant religions in Nigeria) is standing on shaky ground. The relationship is highly characterized by mutual suspicion, mistrust and distrust. In understanding this characterized reality in their interactions, this work interrogated the ambivalence roles played by religious leaders in Nigeria. And in doing this, the ethnographic research method was adopted. As part of its findings, it was discovered that there is a high level of intolerance among Christians and Muslims in Nigeria occasioned by unguarded utterances and abuse of freedom of speech by many uncensored religious leaders. Thus, setting the stage for avoidable and constant religious confrontations among the adherents of the two religious communities in Nigeria. The study recommends that peaceful co-existence can be possible if the government is responsible and responsive enough to address the basic needs of her masses which would reduce largely the manipulation of religion by clerics for personal gain. Furthermore, the place of meaningful dialogue should be embraced by religious leaders across different religious divides. Keywords: Christian-Muslim Relations, Dialogue, Peaceful Co-existence, Religious leaders, Religious Understanding

scholarly journals Scalable Blocking for Very Large Databases

2020 ◽  
pp. 303-319
Author(s):  
Andrew Borthwick ◽  
Stephen Ash ◽  
Bin Pang ◽  
Shehzad Qureshi ◽  
Timothy Jones
Keyword(s):  
Large Databases ◽  
Very Large Databases

Visual Data Mining for Discovering Association Rules

2008 ◽  
pp. 2105-2120
Author(s):  
Kesaraporn Techapichetvanich ◽  
Amitava Datta
Keyword(s):  
Data Mining ◽  
Association Rules ◽  
Association Rule ◽  
Large Data ◽  
Data Sets ◽  
Visual Data Mining ◽  
Useful Knowledge ◽  
Large Databases ◽  
A New Technique ◽  
Mining Association Rule

Both visualization and data mining have become important tools in discovering hidden relationships in large data sets, and in extracting useful knowledge and information from large databases. Even though many algorithms for mining association rules have been researched extensively in the past decade, they do not incorporate users in the association-rule mining process. Most of these algorithms generate a large number of association rules, some of which are not practically interesting. This chapter presents a new technique that integrates visualization into the mining association rule process. Users can apply their knowledge and be involved in finding interesting association rules through interactive visualization, after obtaining visual feedback as the algorithm generates association rules. In addition, the users gain insight and deeper understanding of their data sets, as well as control over mining meaningful association rules.

Relevance Feedback as New Tool for Computer-Aided Diagnosis in Image Databases

2012 ◽  
pp. 86-106
Author(s):  
Issam El Naqa ◽  
Jung Hun Oh ◽  
Yongyi Yang
Keyword(s):  
Relevance Feedback ◽  
Clinical Decision Making ◽  
Clinical Decision ◽  
Computer Aided Diagnosis ◽  
Image Features ◽  
Query Process ◽  
Large Databases ◽  
Computer Aided ◽  
High Level ◽  
Aided Diagnosis

With the ever-growing volume of images used in medicine, the capability to retrieve relevant images from large databases is becoming increasingly important. Despite the recent progress made in the field, its applications in Computer-Aided Diagnosis (CAD) thus far have been limited by the ability to determine the intrinsic mapping between high-level user perception and the underlying low-level image features. Relevance Feedback (RFB) is a post-query process to refine the search by using positive and/or negative indications from the user about the relevance of retrieved images, which has been applied successfully in traditional text-retrieval systems for improving the results of a retrieval strategy. In this chapter, the authors review some recent advances in RFB technology, and discuss its expanding role in content-based image retrieval from medical archives. They provide working examples, based on their experience, for developing machine-learning methods for RFB in mammography and highlight the potential opportunities in this field for CAD applications and clinical decision-making.

Sign in / Sign up

Export Citation Format

Share Document