The price demand relation is a fundamental concept that models how price affects the sale of a product. It is critical to have an accurate estimate of its parameters, as it will impact the company's revenue. The learning has to be performed very efficiently using a small window of a few test points, because of the rapid changes in price demand parameters due to seasonality and fluctuations. However, there are conflicting goals when seeking the two objectives of revenue maximization and demand learning, known as the learn/earn trade-off. This is akin to the exploration/exploitation trade-off that we encounter in machine learning and optimization algorithms. In this paper, we consider the problem of price demand function estimation, taking into account its exploration-exploitation characteristic. We design a new objective function that combines both aspects. This objective function is essentially the revenue minus a term that measures the error in parameter estimates. Recursive algorithms that optimize this objective function are derived. The proposed method outperforms other existing approaches.

The problem of object detection in image and video has been treated by a large number of researchers. Many design factors degrade the reliability of the problem solutions, such as manual modeling of the object, manual features selection, handcrafting architecture, and learning algorithm selection. Here, a generalized object detection and localization system is presented. It has the ability to learn the object model with the processes of feature selection and architecture building automated by adopting the AdaBoost algorithm as a feature selection and meta-learning algorithm. The output of the training phase is a cascade of classifiers which can be used to classify parts of an image within a search window as either object or non object.

Partially runtime-reconfigurable devices allow tasks to be placed and removed dynamically at runtime. For real-time systems, tasks have to complete their work and also to meet their deadlines. It is important to decide at arrival time whether the real-time task could meet its deadline or not. Acceptance tests are concerned with determining whether the incoming task can meet its deadline or not. The utilization bound-based acceptance tests (UBTs) -that accept new tasks till certain utilization limit- were proposed to handle single processor with aperiodic tasks which do not apply to the reconfigurable environment. The rejection ratio was used as a measure for performance of acceptance tests when all the rejected tasks have the same failure cost. However, when acceptance tests are used, the tasks that are not accepted to run are diverted to other system resources and not actually rejected. In this paper, we modify the utilization bound acceptance test to cope with the reconfigurable platform. Although this test requires simple calculations, it may reject tasks that could have been accepted if they wait in the system. Then, we present an exact acceptance test (EAT) for real-time non-preemptive tasks. This test decides exactly whether the incoming task can meet its deadline or not, at its arrival time. The test depends on a look-ahead placement (LAP) strategy. Finally, we propose a new factor, acceptance ratio of workload (AROW), to deal with systems that deploy acceptance tests. The AROW is suitable to measure the performance of acceptance tests as it takes into account sizes and computation times of accepted and diverted tasks. The increase in this ratio means an increase in the work done by the accepted tasks and vice versa. We compare the LAP strategy to the UBT and show its performance regarding the diversion ratio and the AROW measure. Our results show that the LAP strategy outperforms the UBT technique by over 80% using the AROW measure and also enhances the diversion ratio by around 40%.

The phenomenal growth rate of voice over IP applications is fueling the demand of bridging the gap between real-time QoS guarantees and QoS in ad hoc networks. Multimedia applications are designed for stable networks, with unstable mobile ad-hoc networks presenting a challenge for such applications. Real-time applications require mechanisms that guarantee bounded delay and delay jitter. This paper proposes a traffic differentiating queue (TDQ) for AODV to provide higher QoS guarantees for real-time interactive applications. The popular object-oriented network simulator NS2 has been used for the cross evaluation of the proposed TDQ-AODV. The modified algorithm provides significant improvement over the original algorithm in the average packet delay and throughput, in which the average modification improvement reaches up to 1.5times. Moreover, jitter is generally reduced and stabilized over the varying number of connections.

Producing efficient and compact code for embedded DSP processors is very important for nowadays faster and smaller size devices. Because such processors have highly irregular data-path, conventional code generation techniques typically result in inefficient code. Embedded software compilers are expected to make use of the Address Generation Unit (AGU); a feature commonly found in modern embedded DSP processors. This helps in generating optimized offset assignments to program variables in memory, and consequently minimize the overhead instructions dedicated for addresses computations. This paper addresses one of the problems of code optimizations; namely Simple Offset Assignment (SOA) problem.

Partial reconfiguration allows parts of the reconfigurable chip area to be configured without affecting the rest of the chip. This allows placement of tasks at run time on the reconfigurable chip. Area management is a very important issue which highly affect the utilization of the chip and hence the performance. This paper focuses on a major aspect of moving running tasks to free space for new incoming tasks (compaction). We study the effect of compacting running tasks to free more contiguous space on the system performance. First, we introduce a straightforward compaction strategy called blind compaction. We use its performance as a reference to measure the performance of other compaction algorithms. Then we propose a two-dimensional compaction algorithm called one-corner compaction. This algorithm runs with respect to one chip corner. We further extend this algorithm to the four corners of the chip and introduce the 4-corner compaction algorithm. Finally, we compare the performance of these algorithms with some existing compaction strategies (Diessel, 2000). The simulation results show improvement in average task allocation time when using the 4-corner compaction algorithm by 15% and in chip utilization by 16% over the blind compaction. These results outperform the existing strategies

Partially reconfigurable field-programmable gate arrays (FPGAs) allow parts of the chip to be configured at run-time where each part could hold an independent task. Online placement of these tasks result in area fragmentation leading to poor utilization of chip resources. In this paper, we propose a new metric for measuring area fragmentation. The new fragmentation metric gives an indication to the continuity of the occupied (or free) space and not the amount of occupied space. We show how this metric can be extended for multi-dimensional structures. We also show how this metric can be computed efficiently at run time. Next we use this measure during online placement of tasks on FPGAs, such that the chip fragmentation is reduced. Our results show improvement of chip utilization when using this fragmentation aware placement method over other placement methods with well known bottom left first fit, and best fit placement strategies. In real time environment, we achieve an improvement in miss ratio when using the fragmentation aware placement over the bottom left placement strategy.

In this paper, a parallel dynamic load balancing algorithm; which offers a good trade-off between the balancing degree and communication costs; is proposed. The proposed algorithm is analyzed, investigated, and implemented. Also, a new mathematical solution for that algorithm is presented. The analytical solution is tested on standard task graphs (STG) as common test-bed benchmarks. All the different parameters concerned with processors, memory, interconnection networks, and task programs are also presented. The new proposed algorithm is compared and evaluated with a traditional dynamic load balancing one. Experimental results are presented by considering a set of criteria for evaluating the overall performance. Simulation and analytical results proved that the proposed algorithm outperforms the chosen one.

Automated medical image processing and analysis offers a powerful tool for medical diagnosis. In this work we tackle the problem of white blood cell shape analysis based on the morphological characteristics of their outer contour and nuclei. The paper presents a set of preprocessing and segmentation algorithms along with a set of features that are able to recognize and classify different categories of normal white blood cells. The system was tested on gray level images obtained from a CCD camera through a microscope and produced a correct classification rate close to 91%.

In this paper, we present a model-based face detection approach for determining the region occupied by upright frontal views of human faces in unconstrained natural scenes without any priori information. The process is composed of one offline and two online steps. The first step is offline and consists of the modeling stage that generates a model describing the "face" and "non-face" patterns by only few clusters. The second step is the pattern matching stage, which matches small patterns cropped from the input image at different positions and scales against the modeled clusters. The third step is the classification stage, which identifies the specified pattern as face or non-face using the nearest neighbor classifier. The main advantage of the proposed approach is the reduced computation complexity with respect to previously reported systems. This is done by an intelligent arbitration strategy that avoids the exhaustive search at different scales. The system is tested on 50 images containing 253 faces. The correct detection rate is approximately 83% while the false detection count is low.

Snakes, or active contours, have been previously used in computer vision applications to locate and identify objects. However, problems associated with initialization, poor convergence to boundary concavities and high computational complexity, have limited their utility. We present a scheme for object localization and classification based on inexact description of the shape. The approach is based on deformable templates and was tested and proven be very robust with respect to changes in object scale, position and orientation as well as noise and local deformations of shape. A coarse-to-fine algorithm was developed to reduce the computational complexity and achieve efficient implementation. Results of applying the algorithm for automatic identification and localization of industrial parts are presented.

In this paper, we present a 3-D wavelet compression technique for image sequences in video conferencing applications. One of the main requirements for such applications is that the delay has to be within some acceptable limit. When applying wavelet decomposition in the temporal direction, we must store a large number of frames so that such a decomposition can be effective. But this translates into a correspondingly large algorithmic delay. A technique to avoid this obstacle will be proposed. The basic idea is to overlap the decomposition with previous frames that have already been transmitted and selectively transmitting only the new wavelet coefficients.

Three-dimensional (3-D) video compression using wavelets decomposition along the temporal axis dictates that a number of video frames must be buffered to allow for the temporal decomposition. Buffering of frames allows for the temporal correlation to be made use of, and the larger the buffer the more effective the decomposition. One problem inherent in such a set up in interactive applications such as video conferencing, is that buffering translates into a corresponding time delay. We show that 3-D coding of such image sequences can be achieved in the true sense of temporal direction decomposition but with much less buffering requirements. For a practical coder, this can be achieved by introducing an approximation to the way the transform coefficients are encoded. Applying wavelet decomposition using some types of filters may introduce edge errors, which become more prominent in short signal segments. We also present a solution to this problem for the Daubechies (1988) family of filters.

Estimating the flows of rivers can have significant economic impact, as this can help in agricultural water management and in protection from water shortages and possible flood damage. The first goal of this paper is to apply neural networks to the problem of forecasting the flow of the River Nile in Egypt. The second goal of the paper is to utilize the time series as a benchmark to compare between several neural-network forecasting methods.We compare between four different methods to preprocess the inputs and outputs, including a novel method proposed here based on the discrete Fourier series. We also compare between three different methods for the multistep ahead forecast problem: the direct method, the recursive method, and the recursive method trained using a backpropagation through time scheme. We also include a theoretical comparison between these three methods. The final comparison is between different methods to perform longer horizon forecast, and that includes ways to partition the problem into the several subproblems of forecasting K steps ahead.

In this paper, an exact unified analytical expression for the transient (and the steady state) behavior of the Aliasing Error Probability (AEP) of signature analysis testing using single-input external- and internal-XOR LFSR is deduced. The expression, contrary to what is known in the literature, uses the leftmost bit of the LFSR.

A system is developed for the compression of video images, with particular emphasis on video conferencing applications. It combines the well-known quadtree decomposition approach and the discrete cosine transform in an adaptively spatial subsampling and motion compensated scheme. A comparison with the ITU-T H.261 standard for low bit rate video coding reveals that it has better SNR, better perceptual quality (less blocking artifacts), less computational load (about four times less), as well as ease of hardware implementation.

The objective of a computer vision system is to outline the objects in a picture and label them with an appropriate interpretation. This paper proposes a new paradigm for a modular computer vision system which is both data directed and knowledge based. The system consists of three different types of units, two of which are associative data memories implemented as relational databases. The short-term memory (STM) contains the raw color picture data and the most current interpretations and deductions about the original scene. The long-term memory (LTM) contains a detailed model of the scene under consideration. A collection of analysis processors, each of which is specialized for a particular task, can communicate with both of these memories. The information in the LTM remains unchanged during the analysis, while the STM is being continually updated and revised by the appropriate processors. The latter may be conceived of as being activated by certain data conditions in the STM, and using the information in both the LTM and STM to alter the status of the STM.