Efficient use of training data in the n-tuple recognition method

A simple technique is presented for improving the robustness of the n-tuple recognition method against inauspicious choices of architectural parameters, guarding against the saturation problem, and improving the utilisation of small data sets. Experiments are reported which confirm that the method significantly improves performance and reduces saturation in character recognition problems.

The n-tuple classifier: too good to ignore

The n-tuple pattern recognition method has been tested using a selection of 11 large data sets from the European Community StatLog project, so that the results could be compared with those reported for the 23 other algorithms the project tested. The results indicate that this ultra-fast memory-based method is a viable competitor with the others, which include optimisation-based neural network algorithms, even though the theory of memory-based neural computing is less highly developed in terms of statistical theory.

Two Bayesian treatments of the n-tuple recognition method

Two probabilistic interpretations of the n-tuple recognition method are put forward in order to allow this technique to be analysed with the same Bayesian methods used in connection with other neural network models. Elementary demonstrations are then given of the use of maximum likelihood and maximum entropy methods for tuning the model parameters and assisting their interpretation. One of the models can be used to illustrate the significance of overlapping n-tuple samples with respect to correlations in the patterns.

Benchmarking the n-tuple classifier with statlog dataset

The n-tuple recognition method was tested on 11 large real-world data sets and its performance compared to 23 other classification algorithms. On 7 of these, the results show no systematic performance gap between the n-tuple method and the others. Evidence was found to support a possible explanation for why the n-tuple method yields poor results for certain datasets. Preliminary empirical results of a study of the confidence interval (the difference between the two highest scores) are also reported. These suggest a counter-intuitive correlation between the confidence interval distribution and the overall classification performance of the system.

Good-turing estimation for the frequentist n-tuple classifier

We present results concerning the application of the Good-Turing (GT) estimation method to the frequentist n-tuple system. We show that the Good-Turing method can, to a certain extent rectify the Zero Frequency Problem by providing, within a formal framework, improved estimates of small tallies. We also show that it leads to better tuple system performance than Maximum Likelihood estimation (MLE). However, preliminary experimental results suggest that replacing zero tallies with an arbitrary constant close to zero before MLE yields better performance than that of GT system.

A theoretical and experimental account of n-tuple classifier performance

The n-tuple recognition method is briefly reviewed, summarizing the main theoretical results. Large-scale experiments carried out on Stat-Log project datasets confirm this method as a viable competitor to more popular methods due to its speed, simplicity, and accuracy on the majority of a wide variety of classification problems. A further investigation into the failure of the method on certain datasets finds the problem to be largely due to a mismatch between the scales which describe generalization and data sparseness.

The n-tuple network: coping with sub-optimality

The subject of this thesis is the n-tuple net.work (RAMnet). The major advantage of RAMnets is their speed and the simplicity with which they can be implemented in parallel hardware. On the other hand, this method is not a universal approximator and the training procedure does not involve the minimisation of a cost function. Hence RAMnets are potentially sub-optimal. It is important to understand the source of this sub-optimality and to develop the analytical tools that allow us to quantify the generalisation cost of using this model for any given data. We view RAMnets as classifiers and function approximators and try to determine how critical their lack of' universality and optimality is. In order to understand better the inherent. restrictions of the model, we review RAMnets showing their relationship to a number of well established general models such as: Associative Memories, Kamerva's Sparse Distributed Memory, Radial Basis Functions, General Regression Networks and Bayesian Classifiers. We then benchmark binary RAMnet. model against 23 other algorithms using real-world data from the StatLog Project. This large scale experimental study indicates that RAMnets are often capable of delivering results which are competitive with those obtained by more sophisticated, computationally expensive rnodels. The Frequency Weighted version is also benchmarked and shown to perform worse than the binary RAMnet for large values of the tuple size n. We demonstrate that the main issues in the Frequency Weighted RAMnets is adequate probability estimation and propose Good-Turing estimates in place of the more commonly used :Maximum Likelihood estimates. Having established the viability of the method numerically, we focus on providillg an analytical framework that allows us to quantify the generalisation cost of RAMnets for a given datasetL. For the classification network we provide a semi-quantitative argument which is based on the notion of Tuple distance. It gives a good indication of whether the network will fail for the given data. A rigorous Bayesian framework with Gaussian process prior assumptions is given for the regression n-tuple net. We show how to calculate the generalisation cost of this net and verify the results numerically for one dimensional noisy interpolation problems. We conclude that the n-tuple method of classification based on memorisation of random features can be a powerful alternative to slower cost driven models. The speed of the method is at the expense of its optimality. RAMnets will fail for certain datasets but the cases when they do so are relatively easy to determine with the analytical tools we provide.

An exploration of the effect of super large n-tuples on single layer RAMnets

N-tuple recognition systems (RAMnets) are normally modeled using a small number of input lines to each RAM, because the address space grows exponentially with the number of inputs. It is impossible to implement an arbitrarily-large address space as physical memory. But given modest amounts of training data, correspondingly modest numbers of bits will be set in that memory. Hash arrays can therefore be used instead of a direct implementation of the required address space. This paper describes some exploratory experiments using the hash array technique to investigate the performance of RAMnets with very large numbers of input lines. An argument is presented which concludes that performance should peak at a relatively small n-tuple size, but the experiments carried out so far contradict this. Further experiments are needed to confirm this unexpected result.

An intelligent approach for deploying applications in WSN's

Wireless sensor networks have been identified as one of the key technologies for the 21st century. They consist of tiny devices with limited processing and power capabilities, called motes that can be deployed in large numbers of useful sensing capabilities. Even though, they are flexible and easy to deploy, there are a number of considerations when it comes to their fault tolerance, conserving energy and re-programmability that need to be addressed before we draw any substantial conclusions about the effectiveness of this technology. In order to overcome their limitations, we propose a middleware solution. The proposed scheme is composed based on two main methods. The first method involves the creation of a flexible communication protocol based on technologies such as Mobile Code/Agents and Linda-like tuple spaces. In this way, every node of the wireless sensor network will produce and process data based on what is the best for it but also for the group that it belongs too. The second method incorporates the above protocol in a middleware that will aim to bridge the gap between the application layer and low level constructs such as the physical layer of the wireless sensor network. A fault tolerant platform for deploying and monitoring applications in real time offers a number of possibilities for the end user giving him in parallel the freedom to experiment with various parameters, in an effort towards the deployed applications running in an energy efficient manner inside the network. The proposed scheme is evaluated through a number of trials aiming to test its merits under real time conditions and to identify its effectiveness against other similar approaches. Finally, parameters which determine the characteristics of the proposed scheme are also examined.