Artificial Intelligence

1711 Submissions

[15] viXra:1711.0370 [pdf] submitted on 2017-11-20 22:14:32

Finding The Next Term Of Any Given Sequence Using Total Similarity & Dissimilarity {Version 3} ISSN 1751-3030

Authors: Ramesh Chandra Bagadi
Comments: 1 Page.

In this research investigation, the author has detailed a novel scheme of finding the next term of any given sequence.
Category: Artificial Intelligence

[14] viXra:1711.0367 [pdf] submitted on 2017-11-21 00:18:32

One Step Evolution Of Any Real Positive Number {Version 2}

Authors: Ramesh Chandra Bagadi
Comments: 2 Pages.

In this research investigation, the author has detailed the Theory Of One Step Evolution Of Any Real Positive Number.
Category: Artificial Intelligence

[13] viXra:1711.0361 [pdf] submitted on 2017-11-20 02:12:39

Finding The Next Term Of Any Given Sequence Using Total Similarity & Dissimilarity. ISSN 1751-3030

Authors: Ramesh Chandra Bagadi
Comments: 1 Page.

In this research investigation, the author has detailed a novel scheme of finding the next term of any given sequence.
Category: Artificial Intelligence

[12] viXra:1711.0360 [pdf] submitted on 2017-11-20 02:43:10

Ontology Engineering for Robotics

Authors: Frank Schröder
Comments: 8 Pages.

Ontologies are a powerfull alternative to reinforcement learning. They store knowledge in a domain-specific language. The best-practice for implementing ontologies is a distributed version control system which is filled manually by programmers.
Category: Artificial Intelligence

[11] viXra:1711.0359 [pdf] submitted on 2017-11-20 05:21:55

Finding The Next Term Of Any Given Sequence Using Total Similarity & Dissimilarity {New} ISSN 1751-3030

Authors: Ramesh Chandra Bagadi
Comments: 1 Page.

In this research investigation, the author has detailed a novel scheme of finding the next term of any given sequence.
Category: Artificial Intelligence

[10] viXra:1711.0292 [pdf] submitted on 2017-11-12 09:29:57

Strengths and Potential of the SP Theory of Intelligence in General, Human-Like Artificial Intelligence

Authors: J Gerard Wolff
Comments: 20 Pages.

This paper first defines "general, human-like artificial intelligence" (GHLAI) in terms of five principles. In the light of the definition, the paper summarises the strengths and potential of the "SP theory of intelligence" and its realisation in the "computer model", outlined in an appendix, in three main areas: the versatility of the SP system in aspects of intelligence; its versatility in the representation of diverse kinds of knowledge; and its potential for the seamless integration of diverse aspects of intelligence and diverse kinds of knowledge, in any combination. There are reasons to believe that a mature version of the SP system may attain full GHLAI in diverse aspects of intelligence and in the representation of diverse kinds of knowledge.
Category: Artificial Intelligence

[9] viXra:1711.0266 [pdf] submitted on 2017-11-11 03:38:23

Revisit Fuzzy Neural Network: Demystifying Batch Normalization and ReLU with Generalized Hamming Network

Authors: Lixin Fan
Comments: 10 Pages. NIPS 2017 publication.

We revisit fuzzy neural network with a cornerstone notion of generalized hamming distance, which provides a novel and theoretically justified framework to re-interpret many useful neural network techniques in terms of fuzzy logic. In particular, we conjecture and empirically illustrate that, the celebrated batch normalization (BN) technique actually adapts the “normalized” bias such that it approximates the rightful bias induced by the generalized hamming distance. Once the due bias is enforced analytically, neither the optimization of bias terms nor the sophisticated batch normalization is needed. Also in the light of generalized hamming distance, the popular rectified linear units (ReLU) can be treated as setting a minimal hamming distance threshold between network inputs and weights. This thresholding scheme, on the one hand, can be improved by introducing double-thresholding on both positive and negative extremes of neuron outputs. On the other hand, ReLUs turn out to be non-essential and can be removed from networks trained for simple tasks like MNIST classification. The proposed generalized hamming network (GHN) as such not only lends itself to rigorous analysis and interpretation within the fuzzy logic theory but also demonstrates fast learning speed, well-controlled behaviour and state-of-the-art performances on a variety of learning tasks.
Category: Artificial Intelligence

[8] viXra:1711.0265 [pdf] replaced on 2017-11-17 16:28:38

Revisit Fuzzy Neural Network: Bridging the Gap Between Fuzzy Logic and Deep Learning

Authors: Lixin Fan
Comments: 76 Pages.

This article aims to establish a concrete and fundamental connection between two important fields in artificial intelligence i.e. deep learning and fuzzy logic. On the one hand, we hope this article will pave the way for fuzzy logic researchers to develop convincing applications and tackle challenging problems which are of interest to machine learning community too. On the other hand, deep learning could benefit from the comparative research by re-examining many trail-and-error heuristics in the lens of fuzzy logic, and consequently, distilling the essential ingredients with rigorous foundations. Based on the new findings reported in [38] and this article, we believe the time is ripe to revisit fuzzy neural network as a crucial bridge between two schools of AI research i.e. symbolic versus connectionist [93] and eventually open the black-box of artificial neural networks.
Category: Artificial Intelligence

[7] viXra:1711.0250 [pdf] submitted on 2017-11-08 06:37:55

Total Intra Similarity And Dissimilarity Measure For The Values Taken By A Parameter Of Concern. {Version 1}. ISSN 1751-3030

Authors: Ramesh Chandra Bagadi
Comments: 3 Pages.

In this research investigation, the author has detailed a novel method of finding the ‘Total Intra Similarity And Dissimilarity Measure For The Values Taken By A Parameter Of Concern’. The advantage of such a measure is that using this measure we can clearly distinguish the contribution of Intra aspect variation and Inter aspect variation when both are bound to occur in a given phenomenon of concern. This measure provides the same advantages as that provided by the popular F-Statistic measure.
Category: Artificial Intelligence

[6] viXra:1711.0241 [pdf] submitted on 2017-11-07 03:26:43

Dysfunktionale Methoden der Robotik

Authors: Frank Schröder
Comments: 8 Pages. German

Bei der Realisierung von Robotik-Projekten kann man eine ganze Menge verkehrt machen. Damit sind nicht nur kalte Lötstellen oder abstürzende Software gemeint, sondern sehr viel grundsätzlichere Dinge spielen eine Rolle. Um Fehler zu vermeiden, muss man sich zunächst einmal mit den Failure-Patterns näher auseinandersetzen, also jenen Entwicklungsmethoden, nach denen man auf gar keinen Fall einen Roboter bauen und wie die Software möglichst nicht funktionieren sollte.
Category: Artificial Intelligence

[5] viXra:1711.0235 [pdf] submitted on 2017-11-06 20:27:28

Not Merely Memorization in Deep Networks: Universal Fitting and Specific Generalization

Authors: Xiuyi Yang
Comments: 7 Pages.

We reinterpret the training of convolutional neural nets(CNNs) with universal classification theorem(UCT). This theory implies any disjoint datasets can be classified by two or more layers of CNNs based on ReLUs and rigid transformation switch units(RTSUs) we propose here, this explains why CNNs could memorize noise and real data. Subsequently, we present another fresh new hypothesis that CNN is insensitive to some variant from input training data example, this variant relates to original training input by generating functions. This hypothesis means CNNs can generalize well even for randomly generated training data and illuminates the paradox Why CNNs fit real and noise data and fail drastically when making predictions for noise data. Our findings suggest the study about generalization theory of CNNs should turn to generating functions instead of traditional statistics machine learning theory based on assumption that the training data and testing data are independent and identically distributed(IID), and apparently IID assumption contradicts our experiments in this paper.We experimentally verify these ideas correspondingly.
Category: Artificial Intelligence

[4] viXra:1711.0226 [pdf] submitted on 2017-11-07 01:52:12

Theory Of Universal Evolution Along Prime Basis (Time Like) ISSN 1751-3030.

Authors: Ramesh Chandra Bagadi
Comments: 2 Pages.

In this research investigation, the author has detailed the Theory Of Evolution.
Category: Artificial Intelligence

[3] viXra:1711.0208 [pdf] submitted on 2017-11-07 02:22:45

Theory Of Universal Evolution Along Prime Basis (Time Like) {Version 2} ISSN 1751-3030.

Authors: Ramesh Chandra Bagadi
Comments: 2 Pages.

In this research investigation, the author has detailed the Theory Of Evolution.
Category: Artificial Intelligence

[2] viXra:1711.0116 [pdf] submitted on 2017-11-02 23:51:41

Dynamic Thresholding For Linear Binary Classifiers. {Version 2} ISSN 1751-3030

Authors: Ramesh Chandra Bagadi
Comments: 3 Pages.

In this research investigation, the author has detailed a novel method of finding the Thresholding for Linear Binary Classifiers.
Category: Artificial Intelligence

[1] viXra:1711.0034 [pdf] submitted on 2017-11-02 06:05:21

Dynamic Thresholding For Linear Binary Classifiers. ISSN 1751-3030

Authors: Ramesh Chandra Bagadi
Comments: 2 Pages.

In this research investigation, the author has detailed a novel method of finding the Thresholding for Linear Binary Classifiers.
Category: Artificial Intelligence