Artificial Intelligence

2104 Submissions

[4] viXra:2104.0145 [pdf] submitted on 2021-04-24 01:23:39

On the Negation Intensity of a Basic Probability Assignment (Bpa)

Authors: Xiangjun Mi, Chongru Huang, Bingyi Kang
Comments: 29 Pages.

How to obtain negation knowledge is a crucial topic, especially in the field of artificial intelligence. Limited work has been done on the negation of a basic probability assignment (BPA), and which has been studied in depth throughout the literature. However, the aspect of the intensity level of negation enforcement has not yet been investigated. Moreover, let us note that the main characteristic of intelligent systems is just the flexibility for the sake of being able to represent knowledge according to each situation. In general, researchers have a tendency to express the need for cognitive range in the negation. Thus, it would seem very useful to find a wide range of negations under intensity levels in a BPA. Based on these ideas, this paper first proposes a new approach of finding a BPA negation and gives a domain of intensity in which the negation is executed, which is called the negation space. Then, we investigate a number of desirable properties and explore their correlation with entropy. Numerical examples show the characteristics of the proposed negation solution. Finally, we validate the efficiency of the proposed method from the point of view of the Dempster-Shafer belief structure.
Category: Artificial Intelligence

[3] viXra:2104.0111 [pdf] submitted on 2021-04-19 07:35:07

A Novel Conflict Management Considering the Optimal Discounting Weights Using the BWM Method in Dempster-Shafer Evidence Theory

Authors: Lingge Zhou, Xiangjun Mi, Chongru Huang, Yanan Li, Bingyi Kang
Comments: 39 Pages.

Dempster-Shafer evidence theory (DST) is an effective tool for data fusion. In this theory, how to handle conflicts between evidences is still a significant and open issue. In this paper, the best-worst method (BWM) is extended to conflict management in DST. Firstly, a way to determine the best and worst basic probability assignment (BPA) is proposed. Secondly, a novel strategy for determining the optimal weights of BPA using the BWM method is developed. Compared to traditional measure-based conflict management methods, the proposed method has three better performances: (1) A consistency ratio is considered for BPA to check the reliability of the comparisons, producing more reliable results. (2) The final fusion result has less uncertainty, which is more conducive to improve the performance of decision making. (3) The number of BPA comparisons performed during operation (in conflict management) is reduced (especially matrix-based). A practical application in motor rotor fault diagnosis is used to illustrate the effectiveness and practicability of the proposed methodology.
Category: Artificial Intelligence

[2] viXra:2104.0069 [pdf] submitted on 2021-04-12 12:15:16

The Laws of AI

Authors: Egger Mielberg
Comments: 14 Pages.

The truly transparent and predictable work of the artificial intelligence being created can significantly improve the quality of human life, as well as its safety. In our opinion, the self-awareness of artificial intelligence is achievable only if it is independent in making any decision. We present three basic laws of artificial intelligence focused primarily on the possibility of their practical implementation.
Category: Artificial Intelligence

[1] viXra:2104.0005 [pdf] submitted on 2021-04-03 21:35:10

Forcasting and Pattern Analysis of Dhaka Stock Market using LSTM and Phrophet Algorithm

Authors: Tanvir Rahman, Rafia Akhter, Kehinde Lawal, Shamim Ahmed Mazumder, Tamanna Afroz, Ataur Rahman
Comments: 3 Pages.

forecasting or predicting stock market price and the trend has been regarded as a challenging task because of its chaotic nature. The stock market is essentially a non-linear, non-parametric, noisy, and deterministically chaotic system because of liquid money, stock adequacy, human behavior, news related to the stock market, gambling, international money rate, and so on. In a country like Bangladesh, it is very difficult to find any prediction of the stock market especially the Dhaka stock market. Because its trends and forecasting depend on various factors. Understanding the pattern of the stock market and predicting their development and changes are research hotspots in academic and financial circles. Because financial data contain complex, incomplete, and fuzzy information, predicting their development trends is an extremely difficult challenge. Fluctuations in financial data depend on a myriad of correlated constantly changing factors. In this paper, financial productprice data are treated as a one-dimensional series generated bythe projection of a chaotic system composed of multiple factors into the time dimension, and the price series is reconstructed using the time series phase-space reconstruction (PSR) method. An RNN-based prediction model is designed based on the PSR method and long and short-term memory networks (LSTMs) for DL and used to predict stock prices and for predicting stock market data trend we use Facebook open-source model prophet The proposed and some other prediction models are used to predict multiple stock indices for different periods. A comparisonof the results shows that the proposed prediction model has a higher prediction accuracy.
Category: Artificial Intelligence