Artificial Intelligence

1804 Submissions

[17] viXra:1804.0334 [pdf] submitted on 2018-04-24 04:00:45

Introduction of Reflex Based Neural Network

Authors: Liang Yi
Comments: 17 Pages.

This paper introduces a new neural network that works quite different from current neural network model. The RBNN model is based on the concept of conditioned reflex, which widely exists in real creatures. In RBNN, all learning procedure are executed by the neural network itself, which makes it not a complex mathematic model but simple enough to be implemented in real brain. In RBNN, information is organized in a clear way, which makes the whole network a white box rather than a black box, so we can teach the network knowledge easily and fast. This paper shows the power of conditioned reflex as a search tool which can be used as states transfer function in state machine. Using combinations of neurons as symbols which plays the role of letters in traditional state machine, a RBNN can be treated as a state machine with small number of memory unit but huge number of letters.
Category: Artificial Intelligence

[16] viXra:1804.0281 [pdf] submitted on 2018-04-19 09:47:02

Perceptual Significance of Kernel Methods for Natural Image Processing

Authors: Vikas Ramachandra, Truong Nguyen
Comments: 6 Pages.

We explore the unifying connection between kernel regression, Volterra series expansion and multiscale signal decomposition using recent results on function estimation for system identification. We show that using any of these techniques for (non-linear) image processing tasks is (approximately) equivalent. Further, we use the relation between wavelets and independent components of natural images. Kernel methods can be shown to be implicit Volterra series expansions, which are well approximated by wavelets. Wavelets are, in turn, well represented by independent components of natural images. Thus, it can be seen that kernel methods are also near optimal in terms of higher order statistical modeling and approximation of (natural) images. This explains the reason for good results often (perceptually) observed with the use of kernel methods for many image processing problems.
Category: Artificial Intelligence

[15] viXra:1804.0280 [pdf] submitted on 2018-04-19 09:53:13

Superresolution Using Perceptually Significant Side Information

Authors: Vikas Ramachandra, Truong Nguyen
Comments: 4 Pages.

We investigate the problem of super-resolution of images in the presence of side information. In some situations, when some information of the original image is available to the sender, it can be embedded into the low resolution images, either in the pixels themselves or in the headers. This information can be later used when required to reconstruct the superresolved image. For this, a novel multiresolution histogram matching based superresolution procedure is outlined. The proposed technique gives better results compared to contemporary resolution enhancement algorithms, and is especially useful for de-blurring text images captured from mobile phone cameras.
Category: Artificial Intelligence

[14] viXra:1804.0278 [pdf] submitted on 2018-04-19 09:59:42

A Distributed Compressive Sampling Approach for Scene Capture Using an Array of Single Pixel Cameras

Authors: Vikas Ramachandra
Comments: 4 Pages.

This paper presents a method of capturing 3D scene information using an array of single pixel cameras. Based on the recent results for distributed compressive sampling, it is shown here that there could be considerable savings in the measurements required to construct the whole scene, when the correlations between the images captured by the individual cameras in the array is exploited. A technique for doing so for an array of cameras separated by translations along one axis only is illustrated.
Category: Artificial Intelligence

[13] viXra:1804.0249 [pdf] submitted on 2018-04-18 03:52:47

Machine Learning Protein Dynamics Data

Authors: George Rajna
Comments: 32 Pages.

At the University of South Florida, researchers are integrating machine learning techniques into their work studying proteins. [21] Bioinformatics professors Anthony Gitter and Casey Greene set out in summer 2016 to write a paper about biomedical applications for deep learning, a hot new artificial intelligence field striving to mimic the neural networks of the human brain. [20] A team of researchers from the University of Muenster in Germany has now demonstrated that this combination is extremely well suited to planning chemical syntheses—so-called retrosyntheses—with unprecedented efficiency. [19] Two physicists at ETH Zurich and the Hebrew University of Jerusalem have developed a novel machine-learning algorithm that analyses large data sets describing a physical system and extract from them the essential information needed to understand the underlying physics. [18] have come up with a novel machine learning method that enables scientists to derive insights from systems of previously intractable complexity in record time. [17] Quantum computers can be made to utilize effects such as quantum coherence and entanglement to accelerate machine learning. [16] Neural networks learn how to carry out certain tasks by analyzing large amounts of data displayed to them. [15] Who is the better experimentalist, a human or a robot? When it comes to exploring synthetic and crystallization conditions for inorganic gigantic molecules, actively learning machines are clearly ahead, as demonstrated by British Scientists in an experiment with polyoxometalates published in the journal Angewandte Chemie. [14] Machine learning algorithms are designed to improve as they encounter more data, making them a versatile technology for understanding large sets of photos such as those accessible from Google Images. Elizabeth Holm, professor of materials science and engineering at Carnegie Mellon University, is leveraging this technology to better understand the enormous number of research images accumulated in the field of materials science. [13]
Category: Artificial Intelligence

[12] viXra:1804.0237 [pdf] submitted on 2018-04-18 11:04:11

Beyond Back Propagation

Authors: Tofara Moyo
Comments: 1 Page.

5 Protea Lane Newton west
Category: Artificial Intelligence

[11] viXra:1804.0197 [pdf] submitted on 2018-04-14 08:00:32

Artificial Intelligence Accelerates Discovery

Authors: George Rajna
Comments: 40 Pages.

The research group took advantage of a system at SLAC's Stanford Synchrotron Radiation Lightsource (SSRL) that combines machine learning—a form of artificial intelligence where computer algorithms glean knowledge from enormous amounts of data—with experiments that quickly make and screen hundreds of sample materials at a time. [23] Researchers at the UCLA Samueli School of Engineering have demonstrated that deep learning, a powerful form of artificial intelligence, can discern and enhance microscopic details in photos taken by smartphones. [22] Such are the big questions behind one of the new projects underway at the MIT-IBM Watson AI Laboratory, a collaboration for research on the frontiers of artificial intelligence. [21] The possibility of cognitive nuclear-spin processing came to Fisher in part through studies performed in the 1980s that reported a remarkable lithium isotope dependence on the behavior of mother rats. [20] And as will be presented today at the 25th annual meeting of the Cognitive Neuroscience Society (CNS), cognitive neuroscientists increasingly are using those emerging artificial networks to enhance their understanding of one of the most elusive intelligence systems, the human brain. [19] U.S. Army Research Laboratory scientists have discovered a way to leverage emerging brain-like computer architectures for an age-old number-theoretic problem known as integer factorization. [18] have come up with a novel machine learning method that enables scientists to derive insights from systems of previously intractable complexity in record time. [17] Quantum computers can be made to utilize effects such as quantum coherence and entanglement to accelerate machine learning. [16] Neural networks learn how to carry out certain tasks by analyzing large amounts of data displayed to them. [15]
Category: Artificial Intelligence

[10] viXra:1804.0155 [pdf] submitted on 2018-04-11 07:17:37

Deep Learning Smartphone Microscope

Authors: George Rajna
Comments: 36 Pages.

Researchers at the UCLA Samueli School of Engineering have demonstrated that deep learning, a powerful form of artificial intelligence, can discern and enhance microscopic details in photos taken by smartphones. [22] Such are the big questions behind one of the new projects underway at the MIT-IBM Watson AI Laboratory, a collaboration for research on the frontiers of artificial intelligence. [21] The possibility of cognitive nuclear-spin processing came to Fisher in part through studies performed in the 1980s that reported a remarkable lithium isotope dependence on the behavior of mother rats. [20] And as will be presented today at the 25th annual meeting of the Cognitive Neuroscience Society (CNS), cognitive neuroscientists increasingly are using those emerging artificial networks to enhance their understanding of one of the most elusive intelligence systems, the human brain. [19] U.S. Army Research Laboratory scientists have discovered a way to leverage emerging brain-like computer architectures for an age-old number-theoretic problem known as integer factorization. [18] have come up with a novel machine learning method that enables scientists to derive insights from systems of previously intractable complexity in record time. [17] Quantum computers can be made to utilize effects such as quantum coherence and entanglement to accelerate machine learning. [16] Neural networks learn how to carry out certain tasks by analyzing large amounts of data displayed to them. [15] Who is the better experimentalist, a human or a robot? When it comes to exploring synthetic and crystallization conditions for inorganic gigantic molecules, actively learning machines are clearly ahead, as demonstrated by British Scientists in an experiment with polyoxometalates published in the journal Angewandte Chemie. [14] Machine learning algorithms are designed to improve as they encounter more data, making them a versatile technology for understanding large sets of photos such as those accessible from Google Images.
Category: Artificial Intelligence

[9] viXra:1804.0149 [pdf] submitted on 2018-04-09 11:59:00

Machine Learning for Gravitational Waves

Authors: George Rajna
Comments: 21 Pages.

A trio of students from the University of Glasgow have developed a sophisticated artificial intelligence which could underpin the next phase of gravitational wave astronomy. [8] Using data from the first-ever gravitational waves detected last year, along with a theoretical analysis, physicists have shown that gravitational waves may oscillate between two different forms called "g" and "f"-type gravitational waves. [7] Astronomy experiments could soon test an idea developed by Albert Einstein almost exactly a century ago, scientists say. [6] It's estimated that 27% of all the matter in the universe is invisible, while everything from PB&J sandwiches to quasars accounts for just 4.9%. But a new theory of gravity proposed by theoretical physicist Erik Verlinde of the University of Amsterdam found out a way to dispense with the pesky stuff. [5] The proposal by the trio though phrased in a way as to suggest it's a solution to the arrow of time problem, is not likely to be addressed as such by the physics community— it's more likely to be considered as yet another theory that works mathematically, yet still can't answer the basic question of what is time. [4] The Weak Interaction transforms an electric charge in the diffraction pattern from one side to the other side, causing an electric dipole momentum change, which violates the CP and Time reversal symmetry. The Neutrino Oscillation of the Weak Interaction shows that it is a General electric dipole change and it is possible to any other temperature dependent entropy and information changing diffraction pattern of atoms, molecules and even complicated biological living structures.
Category: Artificial Intelligence

[8] viXra:1804.0114 [pdf] submitted on 2018-04-07 10:24:50

Rethinking BICA’s R&D Challenges: Grief Revelations of an Upset Revisionist

Authors: Emanuel Diamant
Comments: 6 Pages.

Biologically Inspired Cognitive Architectures (BICA) is a subfield of Artificial Intelligence aimed at creating machines that emulate human cognitive abilities. What distinguish BICA from other AI approaches is that it based on principles drawn from biology and neuroscience. There is a widespread conviction that nature has a solution for almost all problems we are faced with today. We have only to pick up the solution and replicate it in our design. However, Nature does not easily give up her secrets. Especially, when it is about human brain deciphering. For that reason, large Brain Research Initiatives have been launched around the world. They will provide us with knowledge about brain workflow activity in neuron assemblies and their interconnections. But what is being “flown” (conveyed) via the interconnections the research programme does not disclose. It is implied that what flows in the interconnections is information. But what is information? – that remains undefined. Having in mind BICA’s interest in the matters, the paper will try to clarify the issues.
Category: Artificial Intelligence

[7] viXra:1804.0113 [pdf] submitted on 2018-04-07 10:28:34

Artificial Neural Networks: a Bio-Inspired Revolution or a Long Lasting Misconception and Self-Delusion

Authors: Emanuel Diamant
Comments: 7 Pages. Rejected by the IJCNN 2018, Rio de Janeiro, July 08-13, 2018.

Ali Rahimi, best paper award recipient at NIPS 2017, labelled the current state of Deep Learning (DL) headway as “alchemy”. Yann LeCun, one of the prominent figures in the DL R&D, was insulted by this expression. However, in his response, LeCun did not claimed that DL designers know how and why their DL systems reach so surprising performances. The possible reason for this cautiousness is: No one knows how and in which way system input data is transformed into semantic information at the system’s output. And this, certainly, has its own reason: No one knows what information is! I dare to offer my humble clarification about this obscure and usually untouchable matter. I hope someone would be ready to line up with me.
Category: Artificial Intelligence

[6] viXra:1804.0112 [pdf] submitted on 2018-04-07 11:12:53

Recurrent Capsule Network for Image Generation

Authors: Srikumar Sastry
Comments: 9 Pages.

We have already seen state-of-the-art image generation techniques with Generative Adversarial Networks (Goodfellow et al. 2014), Variational Autoencoder and Recurrent Network for Image generation (K. Gregor et al. 2015). But all these architectures fail to learn object location and pose in images. In this paper, I propose Recurrent Capsule Network based on variational auto encoding framework which can not only preserve equivariance in images in the latent space but also can be used for image classification and generation. For image classification, it can recognise highly overlapping objects due to the use of capsules (Hinton et al. 2011), considerably better than convolutional networks. It can generate images which can be difficult to differentiate from the real data.
Category: Artificial Intelligence

[5] viXra:1804.0094 [pdf] submitted on 2018-04-06 04:33:26

Automated Classification of Hand-Grip Action on Objects Using Machine Learning

Authors: Anju Mishra, Amity University Uttar Pradesh SHANU SHARMA, Amity University Uttar Pradesh SANJAY KUMAR, Oxford Brooks University PRIYA RANJAN, Amity University Uttar Pradesh AMIT UJLAYAN, Gautam Buddha University
Comments: 10 Pages. This is a preprint of a paper under possible publication consideration.

Brain computer interface is the current area of research to provide assistance to disabled persons. To cope up with the growing needs of BCI applications, this paper presents an automated classification scheme for handgrip actions on objects by using Electroencephalography (EEG) data. The presented approach focuses on investigation of classifying correct and incorrect handgrip responses for objects by using EEG recorded patterns. The method starts with preprocessing of data, followed by extraction of relevant features from the epoch data in the form of discrete wavelet transform (DWT), and entropy measures. After computing feature vectors, artificial neural network classifiers used to classify the patterns into correct and incorrect handgrips on different objects. The proposed method was tested on real dataset, which contains EEG recordings from 14 persons. The results showed that the proposed approach is effective and may be useful to develop a variety of BCI based devices to control hand movements. KEYWORDS EEG, Brain computer interface, Machine learning, Hand action recognition
Category: Artificial Intelligence

[4] viXra:1804.0074 [pdf] submitted on 2018-04-04 07:28:11

Biomedical Applications for Deep Learning

Authors: George Rajna
Comments: 30 Pages.

Bioinformatics professors Anthony Gitter and Casey Greene set out in summer 2016 to write a paper about biomedical applications for deep learning, a hot new artificial intelligence field striving to mimic the neural networks of the human brain. [20] A team of researchers from the University of Muenster in Germany has now demonstrated that this combination is extremely well suited to planning chemical syntheses—so-called retrosyntheses—with unprecedented efficiency. [19] Two physicists at ETH Zurich and the Hebrew University of Jerusalem have developed a novel machine-learning algorithm that analyses large data sets describing a physical system and extract from them the essential information needed to understand the underlying physics. [18] Now researchers at the Department of Energy's Lawrence Berkeley National Laboratory (Berkeley Lab) and UC Berkeley have come up with a novel machine learning method that enables scientists to derive insights from systems of previously intractable complexity in record time. [17] Quantum computers can be made to utilize effects such as quantum coherence and entanglement to accelerate machine learning. [16] Neural networks learn how to carry out certain tasks by analyzing large amounts of data displayed to them. [15] Who is the better experimentalist, a human or a robot? When it comes to exploring synthetic and crystallization conditions for inorganic gigantic molecules, actively learning machines are clearly ahead, as demonstrated by British Scientists in an experiment with polyoxometalates published in the journal Angewandte Chemie. [14]
Category: Artificial Intelligence

[3] viXra:1804.0056 [pdf] submitted on 2018-04-05 08:26:14

Computer Recognize Dynamic Events

Authors: George Rajna
Comments: 35 Pages.

Such are the big questions behind one of the new projects underway at the MIT-IBM Watson AI Laboratory, a collaboration for research on the frontiers of artificial intelligence. [21] The possibility of cognitive nuclear-spin processing came to Fisher in part through studies performed in the 1980s that reported a remarkable lithium isotope dependence on the behavior of mother rats. [20] And as will be presented today at the 25th annual meeting of the Cognitive Neuroscience Society (CNS), cognitive neuroscientists increasingly are using those emerging artificial networks to enhance their understanding of one of the most elusive intelligence systems, the human brain. [19] U.S. Army Research Laboratory scientists have discovered a way to leverage emerging brain-like computer architectures for an age-old number-theoretic problem known as integer factorization. [18] have come up with a novel machine learning method that enables scientists to derive insights from systems of previously intractable complexity in record time. [17] Quantum computers can be made to utilize effects such as quantum coherence and entanglement to accelerate machine learning. [16] Neural networks learn how to carry out certain tasks by analyzing large amounts of data displayed to them. [15] Who is the better experimentalist, a human or a robot? When it comes to exploring synthetic and crystallization conditions for inorganic gigantic molecules, actively learning machines are clearly ahead, as demonstrated by British Scientists in an experiment with polyoxometalates published in the journal Angewandte Chemie. [14] Machine learning algorithms are designed to improve as they encounter more data, making them a versatile technology for understanding large sets of photos such as those accessible from Google Images. Elizabeth Holm, professor of materials science and engineering at Carnegie Mellon University, is leveraging this technology to better understand the enormous number of research images accumulated in the field of materials science. [13]
Category: Artificial Intelligence

[2] viXra:1804.0048 [pdf] submitted on 2018-04-03 10:59:48

Machine Learning to Microbial Relationship

Authors: George Rajna
Comments: 30 Pages.

Marculescu, along with ECE Ph.D. student Chieh Lo, has developed a machine learning algorithm—called MPLasso—that uses data to infer associations and interactions between microbes in the GI microbiome. [20] A team of researchers from the University of Muenster in Germany has now demonstrated that this combination is extremely well suited to planning chemical syntheses—so-called retrosyntheses—with unprecedented efficiency. [19] Two physicists at ETH Zurich and the Hebrew University of Jerusalem have developed a novel machine-learning algorithm that analyses large data sets describing a physical system and extract from them the essential information needed to understand the underlying physics. [18] Now researchers at the Department of Energy's Lawrence Berkeley National Laboratory (Berkeley Lab) and UC Berkeley have come up with a novel machine learning method that enables scientists to derive insights from systems of previously intractable complexity in record time. [17] Quantum computers can be made to utilize effects such as quantum coherence and entanglement to accelerate machine learning. [16] Neural networks learn how to carry out certain tasks by analyzing large amounts of data displayed to them. [15]
Category: Artificial Intelligence

[1] viXra:1804.0020 [pdf] submitted on 2018-04-02 03:54:11

A New Neural Network for Artificial General Intelligence

Authors: Haisong Liang
Comments: 44 Pages. Include both English and Chinese version.

Since artificial intelligence was first introduced several decades ago, neural network has achieved remarkable results as one of the most important research methods, and variety of neural network models have been proposed. Usually, for a specific task, we train the network with large amounts of data to develop a mathematical model, making the model produce the expected outputs according to inputs, which also results in the black box problem. In this case, if we study from the perspective of information meanings and their causal relations with the following measures: denote information by neurons; store their relations with links; give neurons a state indicating the strength of information, which can be updated by a state function or input signal; then we can store different information and their relations and control related information's expression with neurons' state. The neural network will become a dynamic system then. More importantly, we can denote different information and logic by designing the topology of neural network and the attributes of the links, and thus having the ability to design and explain every detail of the network precisely, turning neural network into a general information storage, expression, control and processing system, which is also commonly referred as "Strong AI".
Category: Artificial Intelligence