| viXra.org > Artificial Intelligence > 2009 |
 |
 |
| viXra.org > Artificial Intelligence > 2009 |
|
|
[6] viXra:2009.0173 [pdf] submitted on 2020-09-25 20:04:51
Authors: Eren Unlu
Comments: 7 Pages.
We have used the FIFA19 video game open dataset of soccer player attributes and the actual list of squads of national teams that participated in World Cup 2018, which almost coincides in time with the game’s release date. With the intended rationale behind that numerous expert game developers should have spent considerable amount of time to assess each individual player’s attributes; we can develop and test data science and machine learning tools to select national soccer teams in an attempt to assist coaches. The work provides detailed explanatory data analysis and state-of-the-art machine learning and interpretability measures.
Category: Artificial Intelligence
[5] viXra:2009.0165 [pdf] submitted on 2020-09-23 13:48:26
Authors: Jixiang Deng, Yong Deng
Comments: 25 Pages.
Dempster-Shafer evidence theory (evidence theory) has been widely used for its great performance of dealing with uncertainty. Based on evidence theory, researchers have presented different methods to combine evidences. Dempster's rule is the most well-known combination method, which has been applied in many fields. However, Dempster's rule may yield counter-intuitive results when evidences are in high conflict. To improve the performance of combining conflicting evidences, in this paper, we present a new evidence combination method based on Pearson correlation coefficient and weighted graph. The proposed method can correctly identify the target with a high accuracy. Besides, the proposed method has a better performance of convergence compared with other combination methods. In addition, the weighted graph generated by the proposed method can directly represent the relation of different evidences, which can help researchers to determine the reliability of every evidence. Moreover, an experiment is expounded to show the efficiency of the proposed method, and the results are analyzed and discussed.
Category: Artificial Intelligence
[4] viXra:2009.0138 [pdf] submitted on 2020-09-19 20:25:44
Authors: Eren Unlu
Comments: 5 Pages.
We present a novel intuitive graphical representation for daily stock prices, which
we refer as RGBSticks, a variation of classical candle sticks. This representation allows the usage of complex deep learning based techniques, such as deep convolutional autoencoders and deep convolutional generative adversarial networks to produce insightful visualizations for market’s past and future states
Category: Artificial Intelligence
[3] viXra:2009.0061 [pdf] submitted on 2020-09-08 08:49:43
Authors: J. Gerard Wolff
Comments: 37 Pages. As of 2020-09-02, this document has been accepted for publication as a chapter in the book Interpretable Articial Intelligence: A Perspective of Granular Computing, to be published by Springer-Verlag and edited by Witold Pedrycz and Shyi-Ming Chen.
This chapter describes how the SP System, meaning the SP Theory of Intelligence, and its realisation as the SP Computer Model, may promote transparency and granularity in AI, and some other areas of application. The chapter describes how transparency in the workings and output of the SP Computer Model may be achieved via three routes: 1) the program provides a very full audit trail for such processes as recognition, reasoning, analysis of language, and so on. There is also an explicit audit trail for the unsupervised learning of new knowledge; 2) knowledge from the system is
likely to be granular and easy for people to understand; and 3) there are seven principles for the organisation of knowledge which are central in the workings of the SP System and also very familiar to people (eg chunking-with-codes, part-whole hierarchies, and class-inclusion hierarchies), and that kind of familiarity in the way knowledge is structured by the system, is likely to be important in the interpretability, explainability, and transparency of that knowledge. Examples from the SP Computer Model are shown throughout the chapter.
Category: Artificial Intelligence
[2] viXra:2009.0018 [pdf] submitted on 2020-09-03 10:37:12
Authors: J. Gerard Wolff
Comments: 31 Pages. This "technical report" is an adjunct to the paper "Problems in AI research ..." and should be treated as an integral part of that paper
This technical report, an adjunct to the paper "Problems in AI research ...", describes some problems in AI research and how the {\em SP System} (meaning the "SP Theory of Intelligence" and its realisation in the "SP Computer Model") may help to solve them. It also contains a fairly detailed outline of the SP System. Most of the problems considered in this report are described by leading researchers in AI in interviews with science writer Martin Ford, and presented in his book "Architects of Intelligence". Problems and their potential solutions that are described in this report are: the need for more emphasis in research on the use of top-down strategies is met by the way SP has been developed entirely within a top-down framework; the risk of accidents with self-driving vehicles may be minimised via the theory of generalisation within the SP System; the need for strong compositionality in the structure of knowledge is met by processes within the SP Computer Model for unsupervised learning and the organisation of knowledge; although commonsense reasoning and commonsense knowledge are challenges for all theories of AI, the SP System has some promising features; the SP programme of research is one of very few working to establishing the key importance of information compression in AI research; Likewise, the SP programme of research is one of relatively few AI-related research programmes attaching much importance to the biological foundations of intelligence; the SP System lends weight to 'localist' (as compared with 'distributed') views of how knowledge is stored in the brain; compared with deep neural networks, the SP System offers much more scope for adaptation and the representation of knowledge; reasons are given for why the important subjects of motivations and emotions have not so far been considered in the SP programme of research. Evidence in this report, and "Problems in AI research ...", suggests that ***the SP System provides a relatively promising foundation for the development of artificial general intelligence***.
Category: Artificial Intelligence
[1] viXra:2009.0012 [pdf] submitted on 2020-09-02 20:05:43
Authors: J. Gerard Wolff
Comments: 31 Pages. Accepted for publication in the journal Complexity
This paper describes problems in AI research and how the SP System (described in sources referenced in the paper) may help to solve them. Most of the problems considered in the paper are described by leading researchers in AI in interviews with science writer Martin Ford, and reported by him in his book "Architects of Intelligence". These problems, each with potential solutions via SP, are: the divide between symbolic and non-symbolic kinds of knowledge and processing, and how the SP System may bridge the divide; the tendency of deep neural networks (DNNs) to make large and unexpected errors in recognition, something that does not happen with the SP System; in most AI research, unsupervised learning is regarded as a challenge, but unsupervised learning is central in how SP learns; in other AI research, generalisation, with under- and over-generalisation is seen as a problem, but it is a problem that has a coherent solution in the SP System; learning usable knowledge from a single exposure or experience is widely regarded as a problem, but it is a problem that is already solved in the SP System; transfer learning (incorporating old knowledge in new) is seen as an unsolved problem, but it is bedrock in how the SP System learns; there is clear potential for the SP System to solve problems that are prevalent in most AI systems: learning that is slow and greedy for large volumes of data and large computational resources; the SP System provides solutions to problems of transparency in DNNs, where it is difficult to interpret stored knowledge and how it is processed; although there have been successes with DNNs in the processing of natural language, the SP System has strengths in the representation and processing of natural languages which appear to be more in accord with how people process natural language, and these strengths in the SP System are well-integrated with other strengths of the system in aspects of intelligence; by contrast with DNNs, SP has strengths and potential in human-like probabilistic reasoning, and these are well integrated with strengths in other aspects of intelligence; unlike most DNNs, the SP System eliminates the problem of catastrophic forgetting (where new learning wipes out old learning); the SP System provides much of the generality across several aspects of AI which is missing from much research in AI. The strengths and potential of the SP System in comparison with alternatives suggest that {\em the SP System provides a relatively promising foundation for the development of artificial general intelligence}.
Category: Artificial Intelligence
| Third-Party Links: |
|