scholarly journals Computability-theoretic learning complexity

2012 ◽  
Vol 370 (1971) ◽  
pp. 3570-3596 ◽  
Author(s):  
John Case ◽  
Timo Kötzing
Keyword(s):  
Learning Theory ◽  
Main Topic ◽  
Learning Criterion ◽  
Trial And Error ◽  
Easy Method ◽  
Robust Learning ◽  
Intrinsic Complexity ◽  
Data Points ◽  
Computable Operators ◽  
Self Learning

Initially discussed are some of Alan Turing's wonderfully profound and influential ideas about mind and mechanism—including regarding their connection to the main topic of the present study, which is within the field of computability-theoretic learning theory. Herein is investigated the part of this field concerned with the algorithmic, trial-and-error inference of eventually correct programs for functions from their data points. As to the main content of this study: in prior papers, beginning with the seminal work by Freivalds et al. in 1995, the notion of intrinsic complexity is used to analyse the learning complexity of sets of functions in a Gold-style learning setting. Herein are pointed out some weaknesses of this notion. Offered is an alternative based on epitomizing sets of functions—sets that are learnable under a given learning criterion, but not under other criteria that are not at least as powerful. To capture the idea of epitomizing sets, new reducibility notions are given based on robust learning (closure of learning under certain sets of computable operators). Various degrees of epitomizing sets are characterized as the sets complete with respect to corresponding reducibility notions! These characterizations also provide an easy method for showing sets to be epitomizers, and they are then employed to prove several sets to be epitomizing. Furthermore, a scheme is provided to generate easily very strong epitomizers for a multitude of learning criteria. These strong epitomizers are the so-called self-learning sets, previously applied by Case & Kötzing in 2010. These strong epitomizers can be easily generated and employed in a myriad of settings to witness with certainty the strict separation in learning power between the criteria so epitomized and other not as powerful criteria!

Design of a Two-Piece Brassiere Cup From its Data Points Toward its Automation

2020 ◽  
Author(s):  
Kotaro Yoshida ◽  
Hidefumi Wakamatsu ◽  
Eiji Morinaga ◽  
Takahiro Kubo
Keyword(s):  
Three Dimensional ◽  
Developable Surface ◽  
Two Dimensional ◽  
Trial And Error ◽  
Design Efficiency ◽  
Developable Surfaces ◽  
Data Points ◽  
Dimensional Shape ◽  
The Given ◽  
Three Dimensional Shape

Abstract A method to design the two-dimensional shapes of patterns of two piece brassiere cup is proposed when its target three-dimensional shape is given as a cloud of its data points. A brassiere cup consists of several patterns and their shapes are designed by repeatedly making a paper cup model and checking its three-dimensional shape. For improvement of design efficiency of brassieres, such trial and error must be reduced. As a cup model for check is made of paper not cloth, it is assumed that the surface of the model is composed of several developable surfaces. When two lines that consist in the developable surface are given, the surface can be determined. Then, the two-piece brassiere cup can be designed by minimizing the error between the surface and given data points. It was mathematically verified that the developable surface calculated by our propose method can reproduce the given data points which is developable surface.

scholarly journals Optimizing the prediction accuracy of load-settlement behavior of single pile using a self-learning data mining approach

2019 ◽  
Vol 258 ◽  
pp. 02010
Author(s):  
Doddy Prayogo ◽  
Yudas Tadeus Teddy Susanto
Keyword(s):  
Data Mining ◽  
Prediction Accuracy ◽  
Soil Layer ◽  
Training Data ◽  
Support Vector ◽  
Data Mining Approach ◽  
Settlement Behavior ◽  
Data Points ◽  
Single Piles ◽  
Self Learning

Pile foundations usually are used when the upper soil layers are soft clay and, hence, unable to support the structures’ loads. Piles are needed to carry these loads deep into the hard soil layer. Therefore, the safety and stability of pile-supported structures depends on the behavior of the piles. Additionally, an accurate prediction of the piles’ behavior is very important to ensure satisfactory performance of the structures. Although many methods in the literature estimate the settlement of the piles both theoretically and experimentally, methods for comprehensively predicting the load-settlement of piles are very limited. This study develops a new data mining approach called self-learning support vector machine (SL-SVM) to predict the load-settlement behavior of single piles. SL-SVM performance is investigated using 446 training data points and 53 test data points of cone penetration test (CPT) data obtained from the previous literature. The actual prediction accuracy is then compared to other prediction methods using three statistical measurements, including mean absolute error (MAE), coefficient of correlation (R), and root mean square error (RMSE). The obtained results show that SL-SVM achieves better accuracy than does LS-SVM and BPNN. This confirms the capability of the proposed data mining method to model the accurate load-settlement behavior of single piles through CPT data. The paper proposes beneficial insights for geotechnical engineers involved in estimating pile behavior.

Development of Multimedia Teaching Software Based on Self-Learning

2014 ◽  
Vol 998-999 ◽  
pp. 1689-1692
Author(s):  
Hua Ming Yu
Keyword(s):  
Learning Theory ◽  
Software Design ◽  
Procedural Knowledge ◽  
Deep Understanding ◽  
Classification Learning ◽  
Specific Design ◽  
Multimedia Teaching ◽  
The Status ◽  
Self Learning ◽  
Knowledge Classification

Some multimedia teaching software based on the content analysis of already existing problem in the design process to understand the current situation of multimedia teaching software design university. A deep understanding of the knowledge classification learning theory is proposed on the basis of the knowledge classification learning theory as the guiding theory of software design can change the status. And put forward the knowledge classification learning specific design methodology under the guidance of the declarative and procedural knowledge of autonomous learning and multimedia teaching software. Finally based on the above specific strategies to achieve the design these kinds of knowledge learning multimedia teaching software.

A Self-Learning Algorithm for Primitive Actions’ Affordances of Nuclear Power Plant Maintenance Robot

2014 ◽  
Vol 701-702 ◽  
pp. 219-222
Author(s):  
Chun Yang Liu ◽  
Jing Wei Zhang ◽  
Xue Feng Zheng ◽  
Xu Yan Tu
Keyword(s):  
Power Plant ◽  
Nuclear Power Plant ◽  
Nuclear Power ◽  
Learning Algorithm ◽  
Trial And Error ◽  
Successful Operation ◽  
Data Set ◽  
Plant Maintenance ◽  
Self Learning ◽  
Selection Of

In this paper, an algorithm concerning the primitive action affordances of learning nuclear power plant maintenance robot is presented. The algorithm generates a random matching data set through a new matching method, which is utilized for the selection of object operation, with the matching rate improved by trial and error, and then the attempt number for a successful operation is reduced. In the end, simulation is conducted to verify the feasibility and correctness of the proposed algorithm.

scholarly journals Spatiotemporal Dynamics of Reward and Punishment Effects Induced by Associative Learning

2018 ◽  
Author(s):  
Huan Wang ◽  
Killian Kleffner ◽  
Patrick L. Carolan ◽  
Mario Liotti
Keyword(s):  
Associative Learning ◽  
Avoidance Learning ◽  
Event Related Potential ◽  
Perceptual Processing ◽  
Early Event ◽  
Learning Criterion ◽  
Incentive Salience ◽  
Learning Effects ◽  
Trial And Error ◽  
Incentive Value

AbstractWhile reward associative learning has been studied extensively across different species, punishment avoidance learning has received far less attention. Of particular interest is how the two types of learning change perceptual processing of the learned stimuli. We designed a task that required participants to learn the association of emotionally neutral images with reward, punishment, and no incentive value outcomes through trial-and-error. During learning, participants received monetary reward, neutral outcomes or avoided punishment by correctly identifying corresponding images. Results showed an early bias in favor of learning reward associations, in the form of higher accuracy and fewer trials needed to reach learning criterion. We subsequently assessed electrophysiological learning effects with a task in which participants viewed the stimuli with no feedback or reinforcement. Critically, we found modulation of two early event-related potential components for reward images: the frontocentral P2 (170 – 230 ms) and the anterior N2/Early Anterior Positivity (N2/EAP; 210 – 310 ms). We suggest that reward associations may change stimuli detection and incentive salience as indexed by P2 and N2/EAP. We also reported, on an exploratory basis, a late negativity with frontopolar distribution enhanced by punishment images.

Uncertain Interval Data EFCM-ID Clustering Algorithm Based on Machine Learning

2019 ◽  
Vol 31 (2) ◽  
pp. 339-347
Author(s):  
Yimin Mao ◽  
Yinping Liu ◽  
Muhammad Asim Khan ◽  
Jiawei Wang ◽  
Dinghui Mao ◽  
...  
Keyword(s):  
Machine Learning ◽  
Uniform Distribution ◽  
Learning Theory ◽  
Clustering Algorithm ◽  
Interval Data ◽  
Membership Degree ◽  
Running Time ◽  
Spacing Distance ◽  
Data Points ◽  
Clustering Problems

In clustering problems based on fuzzy c-means (FCM) for uncertain interval data, points within the interval are usually assumed to have uniform distribution, resulting in the difficulty of accurately describing the interval. Furthermore, the clustering results are considerably affected by the initial clustering centers, and the update speed of the membership degree is slow. To address these problems, a new clustering algorithm called uncertain FCM for interval data (EFCM-ID) is presented. On the basis of a quartile, a median quartile-spacing distance measurement for generally distributed interval data based on machine learning is designed to precisely determine these data. Simultaneously, we sample the whole dataset and consider the density centers as the initial clustering centers to increase accuracy. We call this method samplingbased density-center selection (SDCS). To reduce the running time, a new measurement based on competitive-learning theory to update the membership is developed. It accelerates the update speed by different degrees according to value of the membership degree. Experiments conducted on synthetic interval datasets show the feasibility of EFCM-ID.

scholarly journals Accelerated design and discovery of perovskites with high conductivity for energy applications through machine learning

2021 ◽  
Vol 7 (1) ◽  
Author(s):  
Pikee Priya ◽  
N. R. Aluru
Keyword(s):  
Machine Learning ◽  
Charge Carriers ◽  
Total Conductivity ◽  
Learning Tools ◽  
Robust Learning ◽  
Energy Applications ◽  
Learning Framework ◽  
Average Ionic Radius ◽  
Data Points ◽  
A Site

AbstractWe use machine learning tools for the design and discovery of ABO3-type perovskite oxides for various energy applications, using over 7000 data points from the literature. We demonstrate a robust learning framework for efficient and accurate prediction of total conductivity of perovskites and their classification based on the type of charge carrier at different conditions of temperature and environment. After evaluating a set of >100 features, we identify average ionic radius, minimum electronegativity, minimum atomic mass, minimum formation energy of oxides for all B-site, and B-site dopant ions of the perovskite as the crucial and relevant predictors for determining conductivity and the type of charge carriers. The models are validated by predicting the conductivity of compounds absent in the training set. We screen 1793 undoped and 95,832 A-site and B-site doped perovskites to report the perovskites with high conductivities, which can be used for different energy applications, depending on the type of the charge carriers.

RELATIVE UNCERTAINTY LEARNING THEORY: AN ESSAY

2004 ◽  
Vol 14 (05) ◽  
pp. 293-311
Author(s):  
SIMONE FIORI
Keyword(s):  
Neural Networks ◽  
Learning Theory ◽  
Uncertainty Theory ◽  
Invariant Manifolds ◽  
Original Learning ◽  
Geometric Analysis ◽  
Learning Criterion ◽  
Relative Uncertainty ◽  
Subspace Analysis ◽  
Matrix Type

The aim of this manuscript is to present a detailed analysis of the algebraic and geometric properties of relative uncertainty theory (RUT) applied to neural networks learning. Through the algebraic analysis of the original learning criterion, it is shown that RUT gives rise to principal-subspace-analysis-type learning equations. Through an algebraic-geometric analysis, the behavior of such matrix-type learning equations is illustrated, with particular emphasis to the existence of certain invariant manifolds.

Mirror representations innate versus determined by experience: A viewpoint from learning theory

2014 ◽  
Vol 37 (2) ◽  
pp. 201-202
Author(s):  
Martin A. Giese
Keyword(s):  
Pattern Recognition ◽  
Genetic Code ◽  
Learning Theory ◽  
Mirror Neurons ◽  
Computational Learning ◽  
Action Perception ◽  
Robust Learning ◽  
Neural Architecture ◽  
Multimodal Representation ◽  
Anatomical Constraints

AbstractFrom the viewpoint of pattern recognition and computational learning, mirror neurons form an interesting multimodal representation that links action perception and planning. While it seems unlikely that all details of such representations are specified by the genetic code, robust learning of such complex representations likely requires an appropriate interplay between plasticity, generalization, and anatomical constraints of the underlying neural architecture.

Sign in / Sign up

Export Citation Format

Share Document