A previous paper (2] showed how to generate a linear discriminant network (LDN) that computes likely faults for a noisy fault detection problem by using a modification of the perceptron learning algorithm called the pocket algorithm. Here we compare the performance of this connectionist model with performance of the optimal Bayesian decision rule for the example that was previously described. W...

In this paper, we propose a new parallel genetic a lge rithm (GA), called Extended Distributed Genetic Algorithm (EDGA), for channel routing problem. The EDGA combines the advantages of previous parallel GA models , viz., master/slave GA model and distributed GA model. In EDGA, the root processor executes the conventional genetic algorithm with global selection on total population and the remai...

Let E and E ′ be elliptic curves over C, with Tate parametrizations p : C ′) sending u ⊗ v to the class of the zero cycle (x, y) − (x, 0) − (0, y) + (0, 0), where x = p(u), y = p ′ (v). We show that, for general u ∈ C * , p * p ′ (u ⊗ (1 − u)) is not zero in CH 2. We also show that the cycle p * p ′ (u ⊗ (1 − u)) is not detectable by a certain class of cohomology theories, including the cohomol...

Appropriate bias is widely viewed as the key to e cient learning and generalization I present a new algorithm the Incremental Delta Bar Delta IDBD algorithm for the learning of appropri ate biases based on previous learning experience The IDBD algorithm is developed for the case of a simple linear learning system the LMS or delta rule with a separate learning rate parameter for each input The I...

are analyzed for extrapolating the 3D information inher­ ent in the scene, the well known correspondence problem We have implemented a new algorithm, called the is ever-present. Magnetic sensors, placed on several places geometric-imprint.s method for extracting key-features of on the body of the human participant, can be used for de­ a given contour. Series of points on the curve are used tect...

paper shows that the ave rage or most likely (optima l) esti­ Many of the processing tasks arising in early vision involve the solution of ill-posed inverse problems. Two techniques that are often used to solve these inverse problems are reg­ ularization and Bayesian modeling. Regularization is used to find a solution that both fits the data and is also suffi­ ciently smooth. Bayesian modeling ...

This paper considers in parallel the scheduling problem for multi class queueing networks and optimization of Markov decision processes It is shown that the value iteration algorithm may perform poorly when the algo rithm is not initialized properly The most typical case where the initial value function is taken to be zero may be a particularly bad choice In contrast if the value iteration algo...

This paper presents a stochastic jump di usion method for optimizing a Bayesian pos terior probability in segmenting range data and their associated re ectance images The algorithm works well on complex real world scenes indoor and outdoor which consist of an unknown number of objects or surfaces of various sizes and types such as planes conics smooth surfaces and cluttered objects like trees a...

Accuracy and speed are the two most im-portant metrics for Network Intrusion Detection or Pre-vention Systems (NIDS/NIPSes). Due to emerging poly-morphic attacks and the fact that in many cases regu-lar expressions (regexes) cannot capture the vulnerabil-ity conditions accurately, the accuracy of existing regex-based NIDS/NIPS systems has become a serious prob-lem. I...

We present a semantic caching approach for optimizing the performance of information mediators A critical problem with informa tion mediators particularly those gathering and integrating information from Web sources is a high query response time This is be cause the data needed to answer user queries is present across several di erent Web sources and in several pages within a source and re trie...