Open Access. Powered by Scholars. Published by Universities.®
- Discipline
-
- Mathematics (465)
- Engineering (59)
- Artificial Intelligence and Robotics (46)
- Applied Mathematics (37)
- Social and Behavioral Sciences (33)
-
- Education (23)
- Electrical and Computer Engineering (23)
- Computer Engineering (19)
- Physics (19)
- Electrical and Electronics (18)
- Life Sciences (17)
- Databases and Information Systems (16)
- Environmental Sciences (15)
- Software Engineering (15)
- Earth Sciences (14)
- Statistics and Probability (14)
- Economics (11)
- Programming Languages and Compilers (10)
- Arts and Humanities (8)
- Mechanical Engineering (8)
- Medicine and Health Sciences (8)
- Civil and Environmental Engineering (7)
- Geography (7)
- Geophysics and Seismology (7)
- Bioinformatics (6)
- Civil Engineering (6)
- Geology (6)
- Biomedical (5)
- Keyword
-
- Technical Reports (390)
- UTEP Computer Science Department (389)
- Machine Learning (30)
- Interval uncertainty (24)
- Fuzzy logic (15)
-
- Optimization (14)
- Deep Learning (13)
- Machine learning (13)
- Interval computations (12)
- Android (8)
- Classification (8)
- Deep learning (8)
- Uncertainty (8)
- Decision making (7)
- Documentation (7)
- Usability (7)
- Data processing (6)
- Game Theory (6)
- Java (6)
- Prosody (6)
- Artificial Intelligence (5)
- Constraints (5)
- Control (5)
- Dialog (5)
- F-transform (5)
- Feasible algorithms (5)
- Functional program verification (5)
- Fuzzy uncertainty (5)
- GPU (5)
- NP-hard (5)
- Publication Year
- Publication
- Publication Type
Articles 361 - 390 of 1174
Full-Text Articles in Computer Sciences
Kinematic Metric Spaces Under Interval Uncertainty: Towards An Adequate Definition, Vladik Kreinovich, Olga Kosheleva, Victor Selivanov
Kinematic Metric Spaces Under Interval Uncertainty: Towards An Adequate Definition, Vladik Kreinovich, Olga Kosheleva, Victor Selivanov
Departmental Technical Reports (CS)
In the physical space, we define distance between the two points as the length of the shortest path connecting these points. Similarly, in space-time, for every pair of events for which the event a can causally effect the event b, we can define the longest proper time t(a,b) over all causal trajectories leading from a to b. The resulting function is known as kinematic metric. In practice, our information about all physical quantities -- including time -- comes from measurement, and measurements are never absolutely precise: the measurement result V is, in general, different from the actual (unknown) value v …
Fuzzy Logic Beyond Traditional "And"- And "Or"-Operations, Vladik Kreinovich, Olga Kosheleva
Fuzzy Logic Beyond Traditional "And"- And "Or"-Operations, Vladik Kreinovich, Olga Kosheleva
Departmental Technical Reports (CS)
In the traditional fuzzy logic, we can use "and"-operations (also known as t-norms) to estimate the expert's degree of confidence in a composite statement A&B based on his/her degrees of confidence d(A) and d(B) in the corresponding basic statements A and B. But what if we want to estimate the degree of confidence in A&B&C in situations when, in addition to the degrees of estimate d(A), d(B), and d(C) of the basic statements, we also know the expert's degrees of confidence in the pairs d(A&B), d(A&C), and d(B&C)? Traditional "and"-operations can provide such an estimate -- but only by ignoring …
Different Concepts, Similar Computational Complexity: Nguyen's Results About Fuzzy And Interval Computations 35 Years Later, Hung T. Nguyen, Vladik Kreinovich
Different Concepts, Similar Computational Complexity: Nguyen's Results About Fuzzy And Interval Computations 35 Years Later, Hung T. Nguyen, Vladik Kreinovich
Departmental Technical Reports (CS)
When we know for sure which values are possible and which are not, we have crisp uncertainty -- of which interval uncertainty is a usual case. In practice, we are often not 100% sure about our knowledge, i.e., we have fuzzy uncertainty -- i.e., we have fuzzy knowledge, of which crisp is a particular case. Usually, general problems are more difficult to solve that most of their particular cases. It was therefore expected that processing fuzzy data is, in general, more computationally difficult than processing interval data -- and indeed, Zadeh's extension principle -- a natural formula for fuzzy computations …
Fault Detection In A Smart Electric Grid: Geometric Analysis, Hector Reyes, Dillon Trinh, Vladik Kreinovich
Fault Detection In A Smart Electric Grid: Geometric Analysis, Hector Reyes, Dillon Trinh, Vladik Kreinovich
Departmental Technical Reports (CS)
The main idea behind a smart grid is to equip the grid with a dense lattice of sensors monitoring the state of the grid. If there is a fault, the sensors closer to the fault will detect larger deviations from the normal readings that sensors that are farther away. In this paper, we show that this fact can be used to locate the fault with high accuracy.
Why Geological Regions?, Daniela Flores, Olga Kosheleva, Vladik Kreinovich
Why Geological Regions?, Daniela Flores, Olga Kosheleva, Vladik Kreinovich
Departmental Technical Reports (CS)
In most practical applications, we approximate the spatial dependence by smooth functions. The main exception is geosciences, where, to describe, e.g., how the density depends on depth and/or on spatial location, geophysicists divide the area into regions on each of which the corresponding quantity is approximately constant. In this paper, we provide a possible explanation for this difference.
Why People Overestimate Small Probabilities?, David Amparan, Vladik Kreinovich
Why People Overestimate Small Probabilities?, David Amparan, Vladik Kreinovich
Departmental Technical Reports (CS)
It is a known empirical fact that people overestimate small probabilities. This fact seems to be inconsistent with the fact that we humans are the product of billions years of improving evolution -- and that we therefore perceive the world as accurately as possible. In this paper, we provide a possible explanation for this seeming contradiction.
Why Rectified Linear Neurons: A Possible Interval-Based Explanation, Jonathan Contreras, Martine Ceberio, Vladik Kreinovich
Why Rectified Linear Neurons: A Possible Interval-Based Explanation, Jonathan Contreras, Martine Ceberio, Vladik Kreinovich
Departmental Technical Reports (CS)
At present, the most efficient machine learning techniques are deep neural networks. In these networks, a signal repeatedly undergoes two types of transformations: linear combination of inputs, and a non-linear transformation of each value v -> s(v). Empirically, the function s(v) = max(v,0) -- known as the rectified linear function -- works the best. There are some partial explanations for this empirical success; however, none of these explanations is fully convincing. In this paper, we analyze this why-question from the viewpoint of uncertainty propagation. We show that reasonable uncertainty-related arguments lead to another possible explanation of why rectified linear functions …
How Probabilistic Methods For Data Fitting Deal With Interval Uncertainty: A More Realistic Analysis, Vladik Kreinovich, Sergey P. Shary
How Probabilistic Methods For Data Fitting Deal With Interval Uncertainty: A More Realistic Analysis, Vladik Kreinovich, Sergey P. Shary
Departmental Technical Reports (CS)
In our previous paper, we showed that a simplified probabilistic approach to interval uncertainty leads to the known notion of a united solution set. In this paper, we show that a more realistic probabilistic analysis of data fitting under interval uncertainty leads to another known notion -- the notion of a tolerable solution set. Thus, the notion of a tolerance solution set also has a clear probabilistic interpretation. Good news is that, in contrast to the united solution set whose computation is, in general, NP-hard, the tolerable solution set can be computed by a feasible algorithm.
Why Neural Networks In The First Place: A Theoretical Explanation, Jonatan Contreras, Martine Ceberio, Olga Kosheleva, Vladik Kreinovich
Why Neural Networks In The First Place: A Theoretical Explanation, Jonatan Contreras, Martine Ceberio, Olga Kosheleva, Vladik Kreinovich
Departmental Technical Reports (CS)
Neural networks -- specifically, deep neural networks -- are, at present, the most effective machine learning techniques. There are reasonable explanations of why deep neural networks work better than traditional "shallow" ones, but the question remains: why neural networks in the first place? why not networks consisting of non-linear functions from some other family of functions? In this paper, we provide a possible theoretical answer to this question: namely, we show that of all families with the smallest possible number of parameters, families corresponding to neurons are indeed optimal -- for all optimality criteria that satisfy some reasonable requirements: : …
Why Daubechies Wavelets Are So Successful, Solymar Ayala Cortez, Laxman Bokati, Aaron Velasco, Vladik Kreinovich
Why Daubechies Wavelets Are So Successful, Solymar Ayala Cortez, Laxman Bokati, Aaron Velasco, Vladik Kreinovich
Departmental Technical Reports (CS)
In many applications, including analysis of seismic signals, Daubechies wavelets perform much better than other families of wavelets. In this paper, we provide a possible theoretical explanation for the empirical success of Daubechies wavelets. Specifically, we show that these wavelets are optimal with respect to any optimality criterion that satisfies the natural properties of scale- and shift-invariance.
Uncertainty: Ideas Behind Neural Networks Lead Us Beyond Kl-Decomposition And Interval Fields, Michael Beer, Olga Kosheleva, Vladik Kreinovich
Uncertainty: Ideas Behind Neural Networks Lead Us Beyond Kl-Decomposition And Interval Fields, Michael Beer, Olga Kosheleva, Vladik Kreinovich
Departmental Technical Reports (CS)
In many practical situations, we know that there is a functional dependence between a quantity q and quantities a1, ..., an, but the exact form of this dependence is only known with uncertainty. In some cases, we only know the class of possible functions describing this dependence. In other cases, we also know the probabilities of different functions from this class -- i.e., we know the corresponding random field or random process. To solve problems related to such a dependence, it is desirable to be able to simulate the corresponding functions, i.e., to have algorithms that transform simple intervals or …
While, In General, Uncertainty Quantification (Uq) Is Np-Hard, Many Practical Uq Problems Can Be Made Feasible, Anderson Gray, Scott Ferson, Olga Kosheleva, Vladik Kreinovich
While, In General, Uncertainty Quantification (Uq) Is Np-Hard, Many Practical Uq Problems Can Be Made Feasible, Anderson Gray, Scott Ferson, Olga Kosheleva, Vladik Kreinovich
Departmental Technical Reports (CS)
In general, many general mathematical formulations of uncertainty quantification problems are NP-hard, meaning that (unless it turned out that P = NP) no feasible algorithm is possible that would always solve these problems. In this paper, we argue that if we restrict ourselves to practical problems, then the correspondingly restricted problems become feasible -- namely, they can be solved by using linear programming techniques.
Ethical Dilemma Of Self-Driving Cars: Conservative Solution, Christian Servin, Vladik Kreinovich, Shahnaz Shahbazova
Ethical Dilemma Of Self-Driving Cars: Conservative Solution, Christian Servin, Vladik Kreinovich, Shahnaz Shahbazova
Departmental Technical Reports (CS)
When designing software for self-driving cars, we need to make an important decision: When a self-driving car encounters an emergency situation in which either the car's passenger or an innocent pedestrian have a good change of being injured or even die, which option should it choose? This has been a subject of many years of ethical discussions -- and these discussions have not yet led to a convincing solution. In this paper, we propose a "conservative" (status quo) solution that does not require making new ethical decisions -- namely, we propose to limit both the risks to passengers and risks …
Localized Learning: A Possible Alternative To Current Deep Learning Techniques, Javier Viana, Kelly Cohen, Anca Ralescu, Stephan Ralescu, Vladik Kreinovich
Localized Learning: A Possible Alternative To Current Deep Learning Techniques, Javier Viana, Kelly Cohen, Anca Ralescu, Stephan Ralescu, Vladik Kreinovich
Departmental Technical Reports (CS)
At present, the most efficient deep learning technique is the use of deep neural networks. However, recent empirical results show that in some situations, it is even more efficient to use "localized" learning -- i.e., to divide the domain of inputs into sub-domains, learn the desired dependence separately on each sub-domain, and then "smooth" the resulting dependencies into a single algorithm. In this paper, we provide theoretical explanation for these empirical successes.
Freedom Of Will, Physics, And Human Intelligence: An Idea, Miroslav Svitek, Vladik Kreinovich, Nguyen Hoang Phuong
Freedom Of Will, Physics, And Human Intelligence: An Idea, Miroslav Svitek, Vladik Kreinovich, Nguyen Hoang Phuong
Departmental Technical Reports (CS)
Among the main fundamental challenges related to physics and human intelligence are: How can we reconcile the free will with the deterministic character of physical equations? What is the physical meaning of extra spatial dimensions needed to make quantum physics consistent? and Why are we often smarter than brain-simulating neural networks? In this paper, we show that while each of these challenges is difficult to resolve on its own, it may be possible to resolve all three of them if we consider them together. The proposed possible solution is that human reasoning uses the extra spatial dimensions. This may sound …
What Is A Reasonable Way To Make Predictions?, Leonardo Orea Amador, Vladik Kreinovich
What Is A Reasonable Way To Make Predictions?, Leonardo Orea Amador, Vladik Kreinovich
Departmental Technical Reports (CS)
Predictions are usually based on what is called laws of nature: many times, we observe the same relation between the states at different moments of time, and we conclude that the same relation will occur in the future. The more times the relation repeats, the more confident we are that the same phenomenon will be re-peated again. This is how Newton's laws and other laws came into being. This is what is called inductive reasoning. However, there are other reasonable approaches. For example, assume that a person speeds and is not caught. This may be repeated two times, three times …
How The Pavement's Lifetime Depends On The Stress Level: An Explanation Of The Empirical Formula, Edgar Daniel Rodriguez Velasquez, Vladik Kreinovich, Olga Kosheleva, Hoang Phuong Nguyen
How The Pavement's Lifetime Depends On The Stress Level: An Explanation Of The Empirical Formula, Edgar Daniel Rodriguez Velasquez, Vladik Kreinovich, Olga Kosheleva, Hoang Phuong Nguyen
Departmental Technical Reports (CS)
We show that natural invariance ideas explain the empirical dependence on the pavement's lifetime on the stress level.
Why Rectified Linear Activation Functions? Why Max-Pooling? A Possible Explanation, Julio C. Urenda, Vladik Kreinovich
Why Rectified Linear Activation Functions? Why Max-Pooling? A Possible Explanation, Julio C. Urenda, Vladik Kreinovich
Departmental Technical Reports (CS)
At present, the most successful machine learning technique is deep learning, that uses rectified linear activation function (ReLU) s(x) = max(x,0) as a non-linear data processing unit. While this selection was guided by general ideas (which were often imprecise), the selection itself was still largely empirical. This leads to a natural question: are these selections indeed the best or are there even better selections? A possible way to answer this question would be to provide a theoretical explanation of why these selections are -- in some reasonable sense -- the best. This paper provides a possible theoretical explanation for this …
Why Normalized Difference Vegetation Index (Ndvi)?, Francisco Zapata, Eric Smith, Vladik Kreinovich, Nguyen Hoang Phuong
Why Normalized Difference Vegetation Index (Ndvi)?, Francisco Zapata, Eric Smith, Vladik Kreinovich, Nguyen Hoang Phuong
Departmental Technical Reports (CS)
Plants play a very important role in ecological systems -- they transform CO2 into oxygen. It is therefore very important to be able to estimate the overall amount of live green vegetation in a given area. The most efficient way to provide such a global analysis is to use remote sensing, i.e., multi-spectral photos taken from satellites, drones, planes, etc. At present, one of the most efficient ways to detect, based on remote sensing data, how much live green vegetation an area contains is to compute the value of the normalized difference vegetation index (NDVI). In this paper, we provide …
Shall We Be Foxes Or Hedgehogs: What Is The Best Balance For Research?, Miroslav Svitek, Olga Kosheleva, Shahnaz Shahbazova, Vladik Kreinovich
Shall We Be Foxes Or Hedgehogs: What Is The Best Balance For Research?, Miroslav Svitek, Olga Kosheleva, Shahnaz Shahbazova, Vladik Kreinovich
Departmental Technical Reports (CS)
Some researchers have few main ideas that they apply to many different problems -- they are called hedgehogs. Other researchers have many ideas but apply them to fewer problems -- they are called foxes. Both approaches have their advantages and disadvantages. What is the best balance between these two approaches? In this paper, we provide general recommendations about this balance. Specifically, we conclude that the optimal productivity is when the time spent on generating new ideas is equal to the time spent on understanding new applications. So, if for a researcher, understanding a new problem is much easier than generating …
As Complexity Rises, Meaningful Statements Lose Precision -- But Why?, Miroslav Svitek, Olga Kosheleva, Vladik Kreinovich
As Complexity Rises, Meaningful Statements Lose Precision -- But Why?, Miroslav Svitek, Olga Kosheleva, Vladik Kreinovich
Departmental Technical Reports (CS)
One of the motivations for Zadeh's development of fuzzy logic -- and one of the explanations for the success of fuzzy techniques -- is the empirical observation that as complexity rises, meaningful statements lose precision. In this paper, we provide a possible explanation for this empirical phenomenon.
Predicting (Economic) Trends: Why Signature Method In Machine Learning, Vladik Kreinovich, Chon Van Le
Predicting (Economic) Trends: Why Signature Method In Machine Learning, Vladik Kreinovich, Chon Van Le
Departmental Technical Reports (CS)
In many practical situations, we can predict the trend -- i.e., how the system will change -- but we cannot predict the exact timing of this change: this timing may depend on many unpredictable factors. For example, we may be sure that the economy will recover, but how fast it will recover may depend on the status of the pandemic, on the weather-affected agriculture input, etc. In such trend predictions, one of the most efficient methods is signature method, which is based on applying machine learning techniques to several special characteristics of the corresponding time series. In this paper, we …
How To Work? How To Study? Shall We Cram For The Exams? And How Is This Related To Life On Earth?, Olga Kosheleva, Vladik Kreinovich, Nguyen Hoang Phuong
How To Work? How To Study? Shall We Cram For The Exams? And How Is This Related To Life On Earth?, Olga Kosheleva, Vladik Kreinovich, Nguyen Hoang Phuong
Departmental Technical Reports (CS)
If we follow the same activity for a long time, our productivity decreases. To increase productivity, a natural idea is therefore to switch to a different activity, and then to switch back and resume the current task. On the other hand, after each switch, we need some time to get back to the original productivity. As a result, too frequent switches are also counterproductive. Natural questions are: shall we switch? if yes, when? In this paper, we use a simple model to provide approximate answers to these questions.
Correcting Interval-Valued Expert Estimates: Empirical Formulas Explained, Laura A. Berrout Ramos, Vladik Kreinovich, Kittawit Autchariyapanitkul
Correcting Interval-Valued Expert Estimates: Empirical Formulas Explained, Laura A. Berrout Ramos, Vladik Kreinovich, Kittawit Autchariyapanitkul
Departmental Technical Reports (CS)
Experts' estimates are approximate. To make decisions based on these estimates, we need to know how accurate these estimate are. Sometimes, experts themselves estimate the accuracy of their estimates -- by providing the interval of possible values instead of a single number. In other cases, we can gauge the accuracy of the experts' estimates by asking several experts to estimates the same quantity and using the interval range of these values. In both situations, sometimes the interval is too narrow -- e.g., if an expert is overconfident. Sometimes, the interval is too wide -- if the expert is too cautious. …
Why Moving Fast And Breaking Things Makes Sense?, Francisco Zapata, Eric Smith, Vladik Kreinovich
Why Moving Fast And Breaking Things Makes Sense?, Francisco Zapata, Eric Smith, Vladik Kreinovich
Departmental Technical Reports (CS)
In the traditional approach to engineering system design, engineers usually come up with several possible designs, each improving on the previous ones. In coming up with these designs, they try their best to make sure that their designs stay within the safety and other constraints, to avoid potential catastrophic crashes. The need for these safety constraints makes this design process reasonably slow. Software engineering at first followed the same pattern, but then realized that since in most cases, failure of a software test does not lead to a catastrophe, it is much faster to first ignore constraints and then adjust …
Why 70/100 Is Satisfactory? Why Five Letter Grades? Why Other Academic Conventions?, Christian Servin, Olga Kosheleva, Vladik Kreinovich
Why 70/100 Is Satisfactory? Why Five Letter Grades? Why Other Academic Conventions?, Christian Servin, Olga Kosheleva, Vladik Kreinovich
Departmental Technical Reports (CS)
Why 70/100 is usually a threshold for a student's satisfactory performance? Why there are usually only five letter grades? Why the usual arrangement of research, teaching, and service is 40-40-20? We show that all these arrangements -- and other similar academic arrangements -- can be explained by two ideas: the Laplace Indeterminacy Principle and the seven plus minus two law.
Blessings, God, Sacrifices: Possible Rational Explanations Of Biblical Ideas, Olga Kosheleva, Vladik Kreinovich
Blessings, God, Sacrifices: Possible Rational Explanations Of Biblical Ideas, Olga Kosheleva, Vladik Kreinovich
Departmental Technical Reports (CS)
In this paper, we show that many seemingly irrational Biblical ideas can actually be rationally interpreted: that God is everywhere, that we can only say what God is not, that God's name is holy, why cannot you bless as many people as you want, etc. We do not insist on our interpretations, there probably are many others, our sole objective was to show that many Biblical ideas can be rationally explained.
How To Gauge Students' Ability To Collaborate?, Christian Servin, Olga Kosheleva, Shahnaz Shahbazova, Vladik Kreinovich
How To Gauge Students' Ability To Collaborate?, Christian Servin, Olga Kosheleva, Shahnaz Shahbazova, Vladik Kreinovich
Departmental Technical Reports (CS)
Usually, we mostly gauge individual students' skills. However, in the modern world, problems are rarely solved by individuals, it is usually a group effort. So, to make sure that students are successful, we also need to gauge their ability to collaborate. In this paper, we describe when it is possible to gauge the students' ability to collaborate; in situations when such a determination is possible, we explain how exactly we can estimate these abilities.
Hardware For Quantized Mixed-Precision Deep Neural Networks, Andres Rios
Hardware For Quantized Mixed-Precision Deep Neural Networks, Andres Rios
Open Access Theses & Dissertations
Recently, there has been a push to perform deep learning (DL) computations on the edge rather than the cloud due to latency, network connectivity, energy consumption, and privacy issues. However, state-of-the-art deep neural networks (DNNs) require vast amounts of computational power, data, and energyâ??resources that are limited on edge devices. This limitation has brought the need to design domain-specific architectures (DSAs) that implement DL-specific hardware optimizations. Traditionally DNNs have run on 32-bit floating-point numbers; however, a body of research has shown that DNNs are surprisingly robust and do not require all 32 bits. Instead, using quantization, networks can run on …
Fast Magnetic Resonance Image Reconstruction With Deep Learning Using An Efficientnet Encoder, Tahsin Rahman
Fast Magnetic Resonance Image Reconstruction With Deep Learning Using An Efficientnet Encoder, Tahsin Rahman
Open Access Theses & Dissertations
This thesis aims to develop an efficient, deep network based method for Magnetic Resonance Imaging (MRI) acceleration through undersampled MR image reconstruction. Deep Neural Networks, particularly Deep Convolutional Networks, have been demonstrated to be highly effective in a wide variety of computer vision tasks, including MRI reconstruction. However, modern highly efficient encoder structures, such as the EfficientNet can potentially reduce reconstruction times further while improving reconstruction quality. To that end, we have developed a multi-channel U-Net MRI reconstruction network which uses an EfficientNet encoder and a custom asymmetric. The network was trained and tested using 5x undersampled multi-channel brain MR …