Multilayer neural networks trained with the back-propagation algorithm constitute the best example of a successful gradient based learning technique. Given an appropriate network architecture, gradient-based learning algorithms can be used to synthesize a complex decision surface that can classify high-dimensional patterns, such as handwritten characters, with minimal preprocessing. This paper reviews various methods applied to handwritten character recognition and compares them on a standard handwritten digit recognition task. Convolutional neural networks, which are specifically designed to deal with the variability of 2D shapes, are shown to outperform all other techniques. Real-life document recognition systems are composed of multiple modules including field extraction, segmentation recognition, and language modeling. A new learning paradigm, called graph transformer networks (GTN), allows such multimodule systems to be trained globally using gradient-based methods so as to minimize an overall performance measure. Two systems for online handwriting recognition are described. Experiments demonstrate the advantage of global training, and the flexibility of graph transformer networks. A graph transformer network for reading a bank cheque is also described. It uses convolutional neural network character recognizers combined with global training techniques to provide record accuracy on business and personal cheques. It is deployed commercially and reads several million cheques per day.

Gradient-based learning applied to document recognition

A new heuristic approach for minimizing possibly nonlinear and non-differentiable continuous space functions is presented. By means of an extensive testbed it is demonstrated that the new method converges faster and with more certainty than many other acclaimed global optimization methods. The new method requires few control variables, is robust, easy to use, and lends itself very well to parallel computation.

/pdf/differential-evolution-a-simple-and-efficient-heuristic-for-4vrb5o5ldc.pdf

Differential Evolution – A Simple and Efficient Heuristic for Global Optimization over Continuous Spaces

Deposits of clastic carbonate-dominated (calciclastic) sedimentary slope systems in the rock record have been identified mostly as linearly-consistent carbonate apron deposits, even though most ancient clastic carbonate slope deposits fit the submarine fan systems better. Calciclastic submarine fans are consequently rarely described and are poorly understood. Subsequently, very little is known especially in mud-dominated calciclastic submarine fan systems. Presented in this study are a sedimentological core and petrographic characterisation of samples from eleven boreholes from the Lower Carboniferous of Bowland Basin (Northwest England) that reveals a >250 m thick calciturbidite complex deposited in a calciclastic submarine fan setting. Seven facies are recognised from core and thin section characterisation and are grouped into three carbonate turbidite sequences. They include: 1) Calciturbidites, comprising mostly of highto low-density, wavy-laminated bioclast-rich facies; 2) low-density densite mudstones which are characterised by planar laminated and unlaminated muddominated facies; and 3) Calcidebrites which are muddy or hyper-concentrated debrisflow deposits occurring as poorly-sorted, chaotic, mud-supported floatstones. These

Phd by thesis

In this paper we present a new data structure for representing Boolean functions and an associated set of manipulation algorithms. Functions are represented by directed, acyclic graphs in a manner similar to the representations introduced by Lee [1] and Akers [2], but with further restrictions on the ordering of decision variables in the graph. Although a function requires, in the worst case, a graph of size exponential in the number of arguments, many of the functions encountered in typical applications have a more reasonable representation. Our algorithms have time complexity proportional to the sizes of the graphs being operated on, and hence are quite efficient as long as the graphs do not grow too large. We present experimental results from applying these algorithms to problems in logic design verification that demonstrate the practicality of our approach.

/pdf/graph-based-algorithms-for-boolean-function-manipulation-2u3osq5jc0.pdf

Graph-Based Algorithms for Boolean Function Manipulation

Our growing dependence on increasingly complex computer and software systems necessitates the development of formalisms, techniques, and tools for assessing functional properties of these systems. One such technique that has emerged in the last twenty years is model checking, which systematically (and automatically) checks whether a model of a given system satisfies a desired property such as deadlock freedom, invariants, and request-response properties. This automated technique for verification and debugging has developed into a mature and widely used approach with many applications. Principles of Model Checking offers a comprehensive introduction to model checking that is not only a text suitable for classroom use but also a valuable reference for researchers and practitioners in the field. The book begins with the basic principles for modeling concurrent and communicating systems, introduces different classes of properties (including safety and liveness), presents the notion of fairness, and provides automata-based algorithms for these properties. It introduces the temporal logics LTL and CTL, compares them, and covers algorithms for verifying these logics, discussing real-time systems as well as systems subject to random phenomena. Separate chapters treat such efficiency-improving techniques as abstraction and symbolic manipulation. The book includes an extensive set of examples (most of which run through several chapters) and a complete set of basic results accompanied by detailed proofs. Each chapter concludes with a summary, bibliographic notes, and an extensive list of exercises of both practical and theoretical nature.

/pdf/principles-of-model-checking-249iimickz.pdf

Principles of Model Checking

Optimized implementation of synchronous models on industrial LTTA systems

Injector characteristics estimation is an essential ingredient for an effective automotive engine control system. An accurate estimation algorithm is presented in the paper. Simulation and experimental results are provided to demonstrate the quality of the algorithm.

http://embedded.eecs.berkeley.edu/Respep/Research/asves/paper1998/Benvenuti_cdc98.pdf

Injector characteristics estimation for spark ignition engines

We present an industrial case study in automotive control of significant complexity: the new common-rail fuel-injection system for Diesel engines developed by Magneti Marelli Powertrain. In this system, an inlet metering valve, inserted before the High Pressure (HP) pump, regulates the fuel flow that supplies the common rail according to the engine operating point (e.g. engine speed and torque). The standard approach to controller design adopted in the automotive in- dustry is based on mean-value models of the plant. However, the fuel injection system exhibits complex discrete-continuous interactions, due to the slow time-varying frequency of the HP pump cycles and the fast sampling frequency of sensing and actuation. Hence, a hybrid approach to controller design is very promizing. In this paper, we present the synthesis of a hybrid multi-rate controller for a Magneti Marelli Powertrain common-rail fuel-injection system for four-cylinder multi-jet engines. Controller design is based on a hybrid model of the injection system. The resulting hybrid controller performs significantly better then previous controllers developed in the company on the basis of mean-value models of the plant.

Hybrid multi-rate control of the common-rail

Residential and commercial buildings in the United States are responsible for up to 73% of the total energy consumption [14]. To reduce the energy consumed by the building stock, both new constructions and existing buildings must be equipped with energy efficient solutions. These solutions are not only architectural, but rely on the use of advanced control algorithms that, based on measurements collected by sensors, compute an optimal control policy and send commands to actuators. Thus, the sensor-actuator network is a key element of building automation systems for energy efficiency. The selection of an optimal network is driven by several concerns including cost. Installation cost is one of the major concerns in building retrofits mainly due to wiring. For new constructions, although wiring is still a concern, the major problem is in making sure that the network is flexible enough to accommodate changes in the building usage that are common over its life-cycle. For these reasons, wireless networks have been always considered as an interesting technology. Wireless sensor networks (WSN) find applications in factory and building automation, environmental monitoring, security systems and in a wide variety of commercial and military systems. However, wireless networks operation cost (which is mainly due to battery replacement) and low reliability have long been the major roadblocks to their adoption. Moreover, the whole-building control policy is often hierarchically structured: room-level controllers communicate to zonelevel controllers, which in turn communicate to floor-level controller and so forth up to the building management system. This structure follows the geometry of the building. At each level, some computational power is required to execute control algorithms which is a challenge in the case of wireless sensor networks. Based on these observations, we argue that the right networked system for building automation is in between a wired and a wireless solution. The first contribution of this work is the selection a design point that seems to be a good compromise between these two extremes. We distinguish between the power network and the data network. The proposed solution is entirely wireless for what concerns the transmission of data. However, we propose a hybrid approach for the power network where sensors are battery powered and are used only for measurement of physical quantities; actuators can be battery powered or not (a power source might be available close to them); and the rest of the components are powered by a wired network. These components provide also some computational power to execute control algorithms. This architecture responds to the needs of being able to reconfigure the sensing platform without major efforts, and to support the structure of a building control system, namely hierarchical and based on building geometry. The second contribution of this paper is a protocol to manage the network so that devices can be easily plugged into the systems, and the network is robust and reliable. The proposed protocol is based on a minimal set of APIs assumed to be provided by the lower layer of the stack. Thus, the proposed protocol can be implemented on other standard protocols such as Zigbee [11]. The third contribution of this paper is to illustrate the implementation details of

/pdf/a-hierarchical-wireless-network-architecture-for-building-307jkv2hhn.pdf

Alberto Sangiovanni-Vincentelli

Papers

Optimized implementation of synchronous models on industrial LTTA systems

Injector characteristics estimation for spark ignition engines

Hybrid multi-rate control of the common-rail

A Hierarchical Wireless Network Architecture for Building Automation and Control Systems

Negative thinking in search methods: application to unate covering