We will introduce a method to extract object boundaries from an image. This method utilizes a deformable curve based on the Self Organizing Map algorithm. The proposed SOM has some unique properties such as batch update and neuron insertion/deletion. These properties can make the SOM converge to object concavities as well as maintain a uniform distribution of neurons along the SOM. In comparison with other traditional active contour methods, this algorithm is less sensitive to initialization more flexible in noisy conditions. It outperforms the Gradient Vector Flow method.

We implement kernel-based principal component analysis to recognize handwritten numbers. Then we analyze the relationship of each numeral type's eigenvalue and eigenvector with its recognition rate. We also present a modified algorithm to improve the robustness as well as the efficiency of the recognition method by employing the secondary training and detection methods from the perspective of nature of kernel function. This method can solve the problem of low recognition rate of a small number of scribbled characters at both low time cost and space complexity. Experiments using 1000 to 5000 test samples all show that our method can achieve 97.8% to 99.0% recognition accuracy.

A dynamic notion of affect (degree of satisfaction) that an agent acquires in competitive iterated game play is developed. Simulated play against both a random environment and a competitor agent is used to study the impact of affect on game playing strategies. For definiteness, the formulation is framed in terms of stock trading. Comments on how affect in game play informs a notion of consciousness along with simulations are given.

Field-programmable gate arrays (FPGAs) are becoming increasingly popular due to low design times, easy testing and implementation procedures and low costs. FPGAs placement and routing are NP-complete problems dealt well with modern tools using heuristic algorithms. As modern FPGAs increase in size and also new capabilities, such as run-time reconfiguration (RTR), are introduced, the complexity of these problems is greatly increased. In this paper we approach both problems using a modified version of Kohonen self-organizing map. The algorithm, consisting of four phases, takes into consideration constraints that may apply to the FPGA design (such as I/O pins, resource constraints like global clock etc). The modified algorithm yields a good topological map of the design to be placed, minimizing the average distance between connecting logic blocks.

The difference between self-organizing maps based phoneme classifiers that emerge for different input languages is studied. For each such language a self-organizing map is trained on mel-frequency cepstral coefficient (MFCC) converted auditory input to form a phoneme classifier. Unsupervised learning is used as the training method. The emerging classes are then compared to the classes found in the International Phonetic Alphabet. Particular class differences across languages and speakers are discussed. We show that SOMs adapt to speakers and languages, even when only given a small training data-set. Additionally, we show that some neurons in SOMs react only to input in one of the two trained languages and that some neurons can be used as word boundary classifiers.

With customer preference databases growing to colossal sizes, collaborative filtering algorithms run into scalability concerns. By reducing the dimensionality of the input space, we ease the demands of predicting users' tastes for films. A movie-to-movie correlation and distance metric are used to decompose the user-movie rating data into two different movie graphs. Using a Kohonen self-organizing map (SOM), the product space can be divided into meaningful clusters centered on the neurons whose weight vectors are nearest the product weights. The SOM-derived clustering is analyzed via a Principal Components Analysis of the data. The clusterings are then evaluated for their effectiveness via quantitative and qualitative observations on the meaningfulness of the groupings of the films.

A new type of model neuron is introduced as a building block of an associative memory. The neuron, which has a number of receptor zones, processes both the amplitude and the frequency of input signals, associating a small number of features encoded by those signals. Using this two-parameter input in our model compared to the one-dimensional inputs of conventional model neurons (e.g., the McCulloch Pitts neuron) offers an increased memory capacity. In our model, there is a competition among inputs in each zone with a subsequent cooperation of the winners to specify the output. The associative memory consists of a network of such neurons. A state-space model is used to define the neurodynamics. We explore properties of the neuron and the network and demonstrate its favorable capacity and recall capabilities. Finally, the network is used in an application designed to find trademarks that sound alike.

A systolic super summer is a cellular apparatus for summing floating-point numbers. The apparatus receives floating-point summands, converting them to fixed-point within a sieve-like cellular array. The emerging fixed-point numbers are summed in a pipelined array of long accumulators. An improved design is presented for the summer's sieve. Although the new sieve is structurally simpler and uses less hardware, the throughput per unit area is the same as that for the previously designed sieve. The new sieve's architectural regularity makes it ideal for implementation in VLSI circuit technology.>

One way to speed up convergence in a large optimization problem is to introduce a smaller, approximate version of the problem at a coarser scale and to alternate between relaxation steps for the fine-scale and coarse-scale problems. Such an optimization method for neural networks governed by quite general objective functions is presented. At the coarse scale, there is a smaller approximating neural net which, like the original net, is nonlinear and has a nonquadratic objective function. The transitions and information flow from fine to coarse scale and back do not disrupt the optimization, and the user need only specify a partition of the original fine-scale variables. Thus, the method can be applied easily to many problems and networks. There is generally about a fivefold improvement in estimated cost under the multiscale method. In the networks to which it was applied, a nontrivial speedup by a constant factor of between two and five was observed, independent of problem size. Further improvements in computational cost are very likely to be available, especially for problem-specific multiscale neural net methods.

A multiscale optimization method for neural networks governed by quite general objective functions is presented. At the coarse scale, there is a smaller, approximating neural net. Like the original net, it is nonlinear and has a nonquadratic objective function, so the coarse-scale net is a more accurate approximation than a quadratic objective would be. The transitions and information flow form fine to coarse scale and back do not disrupt the optimization. The problem need not involve any geometric domain; all that is required is a partition of the original fine-scale variables. Given this partition, the rest of the multiscale optimization method requires no problem-specific design effort on the part of the user, since the mapping between coarse and fine scales is determined. Thus, the method can be applied easily to many problems and networks. Positive experimental results including cost comparisons are shown

A principal limitation in accuracy for scientific computation performed with floating-point arithmetic is due to the computation of repeated sums, such as those that arise in inner products. A systolic super summer of cellular design is proposed for the high-throughput performance of repeated sums of floating-point numbers. The apparatus receives pipelined inputs of streams of summands from one or many sources. The floating-point summands are converted into a fixed-point form by a sieve-like pipelined cellular packet-switching device with signal combining. The emerging fixed-point numbers are then summed in a corresponding network of extremely long accumulators (i.e., super accumulators). At the cell level, the design uses a synchronous model of VLSI. The amount of time the apparatus needs to compute an entire sum depends on the values of summands; at this architectural level, the design is asynchronous. The throughput per unit area of hardware approaches that of a tree network, but without the long wire and signal propagation delay that are intrinsic to tree networks.>

class of search algorithms for locahng the root of a function is devised and studied.Each member of the class of algorithms is globally convergent as m the classical method of binary search However, the class contains members of higher order of convergence as well.These higher order methods require the existence of bounds on derivatives of the function whose root is sought and they proceed by exploiting these bounds.