Probability Distributions.- Linear Models for Regression.- Linear Models for Classification.- Neural Networks.- Kernel Methods.- Sparse Kernel Machines.- Graphical Models.- Mixture Models and EM.- Approximate Inference.- Sampling Methods.- Continuous Latent Variables.- Sequential Data.- Combining Models.

Pattern Recognition and Machine Learning

https://users.encs.concordia.ca/~ymzhang/publications/ARC32-2-98Zhang_pp.229-252.pdf

Bibliographical review on reconfigurable fault-tolerant control systems

This ebook is the first authorized digital version of Kernighan and Ritchie's 1988 classic, The C Programming Language (2nd Ed.). One of the best-selling programming books published in the last fifty years, "K&R" has been called everything from the "bible" to "a landmark in computer science" and it has influenced generations of programmers. Available now for all leading ebook platforms, this concise and beautifully written text is a "must-have" reference for every serious programmers digital library.

As modestly described by the authors in the Preface to the First Edition, this "is not an introductory programming manual; it assumes some familiarity with basic programming concepts like variables, assignment statements, loops, and functions. Nonetheless, a novice programmer should be able to read along and pick up the language, although access to a more knowledgeable colleague will help."

The C programming language

With advanced computer technology, systems are getting larger to fulfill more complicated tasks: however, they are also becoming less reliable. Consequently, testing is an indispensable part of system design and implementation; yet it has proved to be a formidable task for complex systems. This motivates the study of testing finite stare machines to ensure the correct functioning of systems and to discover aspects of their behavior. A finite state machine contains a finite number of states and produces outputs on state transitions after receiving inputs. Finite state machines are widely used to model systems in diverse areas, including sequential circuits, certain types of programs, and, more recently, communication protocols. In a testing problem we have a machine about which we lack some information; we would like to deduce this information by providing a sequence of inputs to the machine and observing the outputs produced. Because of its practical importance and theoretical interest, the problem of testing finite state machines has been studied in different areas and at various times. The earliest published literature on this topic dates back to the 1950's. Activities in the 1960's mid early 1970's were motivated mainly by automata theory and sequential circuit testing. The area seemed to have mostly died down until a few years ago when the testing problem was resurrected and is now being studied anew due to its applications to conformance testing of communication protocols. While some old problems which had been open for decades were resolved recently, new concepts and more intriguing problems from new applications emerge. We review the fundamental problems in testing finite state machines and techniques for solving these problems, tracing progress in the area from its inception to the present and the stare of the art. In addition, we discuss extensions of finite state machines and some other topics related to testing.

Principles and methods of testing finite state machines-a survey

Ultrathin films of single-walled carbon nanotubes (SWNTs) represent an attractive, emerging class of material, with properties that can approach the exceptional electrical, mechanical, and optical characteristics of individual SWNTs, in a format that, unlike isolated tubes, is readily suitable for scalable integration into devices. These features suggest the potential for realistic applications as conducting or semiconducting layers in diverse types of electronic, optoelectronic and sensor systems. This article reviews recent advances in assembly techniques for forming such films, modeling and experimental work that reveals their collective properties, and engineering aspects of implementation in sensors and in electronic devices and circuits with various levels of complexity. A concluding discussion provides some perspectives on possibilities for future work in fundamental and applied aspects.

http://www.ifc.dicp.ac.cn/library/cailiao/pdf1/11.pdf

Ultrathin Films of Single-Walled Carbon Nanotubes for Electronics and Sensors: A Review of Fundamental and Applied Aspects

Guest Editorial: Special Section on "To be Safe and Dependable in the Era of Artificial Intelligence: Emerging Techniques for Trusted and Reliable Machine Learning"

The recent development of nanoscale materials and assembly techniques has resulting in the manufacturing of high-density computational systems. These systems consist of nanometer-scale elements and are likely to have many manufacturing imperfections (defects); thus, defect-tolerance is considered as one of the most some algorithms for repairing defective crosspoints in a nanoscale crossbar architecture; furthermore we estimate the efficiency and cost-effectiveness of each algorithm. Also, for a given design and manufacturing environment, we propose a cost-driven method to find a balanced solution by which figures of merit such as area, repair time and reconfiguration cost can be taken into account. Probabilistic parameters are utilized in the proposed cost-driven method for added flexibility.

/pdf/cost-driven-repair-of-a-nanowire-crossbar-architecture-be8ivwtk1y.pdf

Cost-Driven Repair of a Nanowire Crossbar Architecture

The complexity of Machine Learning (ML) systems increases each year, with current implementations of large language models or text-to-image generators having billions of parameters and requiring billions of arithmetic operations. As these systems are widely utilized, ensuring their reliable operation is becoming a design requirement. Traditional error detection mechanisms introduce circuit or time redundancy that significantly impacts system performance. An alternative is the use of Concurrent Error Detection (CED) schemes that operate in parallel with the system and exploit their properties to detect errors. CED is attractive for large ML systems because it can potentially reduce the cost of error detection. In this paper, we introduce Concurrent Classifier Error Detection (CCED), a scheme to implement CED in ML systems using a concurrent ML classifier to detect errors. CCED identifies a set of check signals in the main ML system and feeds them to the concurrent ML classifier that is trained to detect errors. The proposed CCED scheme has been implemented and evaluated on two widely used large-scale ML models: Contrastive Language Image Pretraining (CLIP) used for image classification and Bidirectional Encoder Representations from Transformers (BERT) used for natural language applications. The results show that more than 95 percent of the errors are detected when using a simple Random Forest classifier that is order of magnitude simpler than CLIP or BERT. These results illustrate the potential of CCED to implement error detection in large-scale ML models.

Concurrent Classifier Error Detection (CCED) in Large Scale Machine Learning Systems

A Look-Up Table (LUT) shows modest performance (delay and power) when used as a universal logic module (ULM) for implementing all possible combinational functions; moreover, the complete programmability of a LUT (so for all functions) incurs in a significant circuit complexity. Few approaches have been proposed by which a LUT is replaced by circuits; this is possible because in practice, the number of inputs (variables) in combinational circuit design is often no more than 4. This paper investigates a partial ULM (PULM) in which only specific functions are incorporated in the logic module. As instances of a PULM, two types of a 4-input PULM are proposed in this paper. These PULMs are initially found by profiling the most common combinational functions in benchmarks to be used in the module circuits. A number of functions of the highest occurrence (4 and 8 functions respectively) are utilized and their performances with respect to average delay and power dissipation are assessed by HSPICE simulation. These simulation results show that the proposed PULMs achieve significant improvements in critical path delay and power consumption; the improvement in critical path delay (power dissipation) decreases (increases) when the number of implemented functions in the module increases. Benchmarks are then utilized for comparative assessment of the proposed PULMs with existing approaches found in the technical literature; also in these cases the PULM-based designs outperform other schemes.

Partially universal modules for high performance logic circuit design

Feature-embedding triplet networks (TNs) with three symmetric subchannels are very promising for similarity-measuring applications. This paper proposes a novel separately constrained triple loss (SCTL) function that applies to TNs for classification. Through minimizing the intra-class distance and maximizing the inter-class distance, SCTL eliminates possible false solutions and provides insight into the dependency of training based on these two terms. Based on this dependency, the strategy of selecting hyperparameters in SCTL is also analyzed to further improve performance. The effectiveness of the proposed SCTL is evaluated based on TNs with multi-layer perceptrons; the results show that compared to all existing loss functions, the use of SCTL offers the best classification accuracy for the TNs, while incurring in negligible hardware overhead (e.g., only a 0.0002% area overhead of the subnetworks).

Fabrizio Lombardi

Papers

Guest Editorial: Special Section on "To be Safe and Dependable in the Era of Artificial Intelligence: Emerging Techniques for Trusted and Reliable Machine Learning"

Cost-Driven Repair of a Nanowire Crossbar Architecture

Concurrent Classifier Error Detection (CCED) in Large Scale Machine Learning Systems

Partially universal modules for high performance logic circuit design

Feature-Embedding Triplet Networks with a Separately Constrained Loss Function