Shlomo Geva

Adaptive nearest neighbor pattern classification

A variant of nearest-neighbor (NN) pattern classification and supervised learning by learning vector quantization (LVQ) is described. The decision surface mapping method (DSM) is a fast supervised learning algorithm and is a member of the LVQ family of algorithms. A relatively small number of prototypes are selected from a training set of correctly classified samples. The training set is then used to adapt these prototypes to map the decision surface separating the classes. This algorithm is compared with NN pattern classification, learning vector quantization, and a two-layer perceptron trained by error backpropagation. When the class boundaries are sharply defined (i.e., no classification error in the training set), the DSM algorithm outperforms these methods with respect to error rates, learning rates, and the number of prototypes required to describe class boundaries.

A constructive method for multivariate function approximation by multilayer perceptrons

Mathematical theorems establish the existence of feedforward multilayered neural networks, based on neurons with sigmoidal transfer functions, that approximate arbitrarily well any continuous multivariate function. However, these theorems do not provide any hint on how to find the network parameters in practice. It is shown how to construct a perceptron with two hidden layers for multivariate function approximation. Such a network can perform function approximation in the same manner as networks based on Gaussian potential functions, by linear combination of local functions.

A cartpole experiment benchmark for trainable controllers

The inverted pendulum problem, i.e., the cartpole, which is often used for demonstrating the success of neural network learning methods, is addressed. It is shown that a random search in weight space can quickly uncover coefficients (weights) for controllers that work over a wide range of initial conditions. This result indicates that success in finding a satisfactory neural controller is not sufficient proof for the effectiveness of unsupervised training methods. By analyzing the dynamics of the linear controller, the cartpole problem is reformulated to make it a more stringent test for neural training methods. A review of the literature on unsupervised training methods for cartpole controllers shows that the published results are difficult to compare and that for most of the methods there is not clear evidence of better performance than the random search method.<<ETX>>

Overview of the INEX 2009 link the wiki track

Wikipedia is becoming ever more popular. Linking between documents is typically provided in similar environments in order to achieve collaborative knowledge sharing. However, this functionality in Wikipedia is not integrated into the document creation process and the quality of automatically generated links has never been quantified. The Link the Wiki (LTW) track at INEX in 2007 aimed at producing a standard procedure, metrics and a discussion forum for the evaluation of link discovery. The tasks offered by the LTW track as well as its evaluation present considerable research challenges. This paper briefly described the LTW task and the procedure of evaluation used at LTW track in 2007. Automated link discovery methods used by participants are outlined. An overview of the evaluation results is concisely presented and further experiments are reported.

Rule Extraction from Local Cluster Neural Nets

This paper describes RULEX, a technique for providing an explanation component for local cluster (LC) neural networks. RULEX extracts symbolic rules from the weights of a trained LC net. LC nets are a special class of multilayer perceptrons that use sigmoid functions to generate localised functions. LC nets are well suited to both function approximation and discrete classification tasks. The restricted LC net is constrained in such a way that the local functions are ‘axis parallel’ thus facilitating rule extraction. This paper presents results for the LC net on a wide variety of benchmark problems and shows that RULEX produces comprehensible, accurate rules that exhibit a high degree of fidelity with the LC network from which they were extracted.

GPX: gardens point XML IR at INEX 2005

The INEX 2005 evaluation consisted of numerous tasks that required different approaches. In this paper we described the approach that we adopted to satisfy the requirements of all the tasks, CAS and CO, in Thorough, Focused, and Fetch Browse mode, using the same underlying system The retrieval approach is based on the construction of a collection sub-tree, consisting of all nodes that contain one or more of the search terms. Nodes containing search terms are then assigned a score using a TF_IDF variant, scores are propagated upwards in the document XML tree, and finally all XML elements are ranked. We present results that demonstrate that the approach is versatile and produces consistently good performance across all INEX 2005 tasks.

Overview of the INEX 2008 Ad Hoc Track

This paper gives an overview of the INEX 2008 Ad Hoc Track. The main goals of the Ad Hoc Track were two-fold. The first goal was to investigate the value of the internal document structure (as provided by the XML mark-up) for retrieving relevant information. This is a continuation of INEX 2007 and, for this reason, the retrieval results are liberalized to arbitrary passages and measures were chosen to fairly compare systems retrieving elements, ranges of elements, and arbitrary passages. The second goal was to compare focused retrieval to article retrieval more directly than in earlier years. For this reason, standard document retrieval rankings have been derived from all runs, and evaluated with standard measures. In addition, a set of queries targeting Wikipedia have been derived from a proxy log, and the runs are also evaluated against the clicked Wikipedia pages. The INEX 2008 Ad Hoc Track featured three tasks: For the Focused Task a ranked-list of non-overlapping results (elements or passages) was needed. For the Relevant in Context Task non-overlapping results (elements or passages) were returned grouped by the article from which they came. For the Best in Context Task a single starting point (element start tag or passage start) for each article was needed. We discuss the results for the three tasks, and examine the relative effectiveness of element and passage retrieval. This is examined in the context of content only (CO, or Keyword) search as well as content and structure (CAS, or structured) search. Finally, we look at the ability of focused retrieval techniques to rank articles, using standard document retrieval techniques, both against the judged topics as well as against queries and clicks from a proxy log.

The effects of impurity elements on the reduction of wustite and magnetite to iron in CO/CO2 and H2/H2O gas mixtures

The reduction of dense wustite and magnetite samples in CO/CO2 and H2/H2O gas mixtures has shown that impurity elements in solid solution in the oxides can significantly affect the reaction mechanisms operative during reduction and the conditions for porous iron growth. In general, the presence of P, Mg, Ti, Si, Ca, K, and Na in wustite favors, in order of increasing strength, the formation of the porous iron product morphology. Aluminum, on the other hand, significantly reduces the range of gas conditions over which the porous iron microstructure may be obtained.

Overview of the INEX 2009 ad hoc track

This paper gives an overview of the INEX 2009 Ad Hoc Track. The main goals of the Ad Hoc Track were three-fold. The first goal was to investigate the impact of the collection scale and markup, by using a new collection that is again based on a the Wikipedia but is over 4 times larger, with longer articles and additional semantic annotations. For this reason the Ad Hoc track tasks stayed unchanged, and the Thorough Task of INEX 2002-2006 returns. The second goal was to study the impact of more verbose queries on retrieval effectiveness, by using the available markup as structural constraints--now using both the Wikipedias layout-based markup, as well as the enriched semantic markup--and by the use of phrases. The third goal was to compare different result granularities by allowing systems to retrieve XML elements, ranges of XML elements, or arbitrary passages of text. This investigates the value of the internal document structure (as provided by the XML mark-up) for retrieving relevant information. The INEX 2009 Ad Hoc Track featured four tasks: For the Thorough Task a ranked-list of results (elements or passages) by estimated relevance was needed. For the Focused Task a ranked-list of non-overlapping results (elements or passages) was needed. For the Relevant in Context Task non-overlapping results (elements or passages) were returned grouped by the article from which they came. For the Best in Context Task a single starting point (element start tag or passage start) for each article was needed. We discuss the setup of the track, and the results for the four tasks.

Local cluster neural net: Architecture, training and applications

Abstract This paper describes the structure, training and computational abilities of the local cluster (LC) artificial neural net architecture. LC nets are a special class of multilayer perceptrons that use sigmoid functions to generate localised functions. LC nets train as fast as radial basis functions nets and are more general. They are well suited for both, multi-dimensional function approximation and discrete classification. The LC net is the result of our search for a widely applicable neural net architecture suitable for low-cost hardware realisation. The LC net seem particularly well suited for analog VLSI realisation of small-size, low-power, fully parallel neural net chip for real time control applications.