An empirical study of Hoelfding Racing for model selction in K-nearest neighbor classification

Racing algorithms have recently been proposed as a general-purpose method for performing model selection in machine teaming algorithms. In this paper, we present an empirical study of the Hoeffding racing algorithm for selecting the k parameter in a simple k-nearest neighbor classifier. Fifteen widely-used classification datasets from UCI are used and experiments conducted across different confidence levels for racing. The results reveal a significant amount of sensitivity of the k-nn classifier to its model parameter value. The Hoeffding racing algorithm also varies widely in its performance, in terms of the computational savings gained over an exhaustive evaluation. While in some cases the savings gained are quite small, the racing algorithm proved to be highly robust to the possibility of erroneously eliminating the optimal models. All results were strongly dependent on the datasets used.

Short-term traffic volume forecasting : a k-nearest neighbor approach enhanced by constrained linearly sewing principle component algorithm

To enhance the performance of the k-nearest neighbors approach in forecasting short-term traffic volume, this paper proposed and tested a two-step approach with the ability of forecasting multiple steps. In selecting k-nearest neighbors, a time constraint window is introduced, and then local minima of the distances between the state vectors are ranked to avoid overlappings among candidates. Moreover, to control extreme values’ undesirable impact, a novel algorithm with attractive analytical features is developed based on the principle component. The enhanced KNN method has been evaluated using the field data, and our comparison analysis shows that it outperformed the competing algorithms in most cases.

Simultaneous feature selection and feature weighting using Hybrid Tabu Search/K-nearest neighbor classifier

Feature selection and feature weighting are useful techniques for improving the classification accuracy of K-nearest-neighbor (K-NN) rule. The term feature selection refers to algorithms that select the best subset of the input feature set. In feature weighting, each feature is multiplied by a weight value proportional to the ability of the feature to distinguish pattern classes. In this paper, a novel hybrid approach is proposed for simultaneous feature selection and feature weighting of K-NN rule based on Tabu Search (TS) heuristic. The proposed TS heuristic in combination with K-NN classifier is compared with several classifiers on various available data sets. The results have indicated a significant improvement in the performance in classification accuracy. The proposed TS heuristic is also compared with various feature selection algorithms. Experiments performed revealed that the proposed hybrid TS heuristic is superior to both simple TS and sequential search algorithms. We also present results for the classification of prostate cancer using multispectral images, an important problem in biomedicine.

Speaker Normalization Methods for Vowel Cognition: Comparative Analysis Using Neural Network and Nearest Neighbor Classifiers

Intrinsic and extrinsic speaker normalization methods are systematically compared using a neural network (fuzzy ARTMAP) and L1 and L2 K-Nearest Neighbor (K-NN) categorizers trained and tested on disjoint sets of speakers of the Peterson-Barney vowel database. Intrinsic methods include one nonscaled, four psychophysical scales (bark, bark with endcorrection, mel, ERB), and three log scales, each tested on four combinations of F0 , F1, F2, F3. Extrinsic methods include four speaker adaptation schemes, each combined with the 32 intrinsic methods: centroid subtraction across all frequencies (CS), centroid subtraction for each frequency (CSi), linear scale (LS), and linear transformation (LT). ARTMAP and KNN show similar trends, with K-NN performing better, but requiring about ten times as much memory. The optimal intrinsic normalization method is bark scale, or bark with endcorrection, using the differences between all frequencies (Diff All). The order of performance for the extrinsic methods is LT, CSi, LS, and CS, with fuzzy ARTMAP performing best using bark scale with Diff All; and K-NN choosing psychophysical measures for all except CSi.

Boosting Nearest Neighbor Classifiers for Multiclass Recognition

This paper introduces an algorithm that uses boosting to learn a distance measure for multiclass k-nearest neighbor classification. Given a family of distance measures as input, AdaBoost is used to learn a weighted distance measure, that is a linear combination of the input measures. The proposed method can be seen both as a novel way to learn a distance measure from data, and as a novel way to apply boosting to multiclass recognition problems, that does not require output codes. In our approach, multiclass recognition of objects is reduced into a single binary recognition task, defined on triples of objects. Preliminary experiments with eight UCI datasets yield no clear winner among our method, boosting using output codes, and k-nn classification using an unoptimized distance measure. Our algorithm did achieve lower error rates in some of the datasets, which indicates that, in some domains, it may lead to better results than existing methods.

Nearest-neighbor Queries in Probabilistic Graphs

Large probabilistic graphs arise in various domains spanning from social networks to biological and communication networks. An important query in these graphs is the k nearest-neighbor query, which involves finding and reporting the k closest nodes to a specific node. This query assumes the existence of a measure of the "proximity" or the "distance" between any two nodes in the graph. To that end, we propose various novel distance functions that extend well known notions of classical graph theory, such as shortest paths and random walks. We argue that many meaningful distance functions are computationally intractable to compute exactly. Thus, in order to process nearest-neighbor queries, we resort to Monte Carlo sampling and exploit novel graph-transformation ideas and pruning opportunities. In our extensive experimental analysis, we explore the trade-offs of our approximation algorithms and demonstrate that they scale well on real-world probabilistic graphs with tens of millions of edges.

Comparison of classification ability of hyperball algorithms to neural network and k-nearest neighbour algorithms

The main focus of this thesis is to evaluate and compare Hyperbalilearning algorithm (HBL) to other learning algorithms. In this work HBL is compared to feed forward artificial neural networks using back propagation learning, K-nearest neighbor and 103 algorithms. In order to evaluate the similarity of these algorithms, we carried out three experiments using nine benchmark data sets from UCI machine learning repository. The first experiment compares HBL to other algorithms when sample size of dataset is changing. The second experiment compares HBL to other algorithms when dimensionality of data changes. The last experiment compares HBL to other algorithms according to the level of agreement to data target values. Our observations in general showed, considering classification accuracy as a measure, HBL is performing as good as most ANn variants. Additionally, we also deduced that HBL.:s classification accuracy outperforms 103's and K-nearest neighbour's for the selected data sets.

Hybrid incremental modeling based on least squares and fuzzy K-NN for monitoring tool wear in turning processes

There is now an emerging need for an efficient modeling strategy to develop a new generation of monitoring systems. One method of approaching the modeling of complex processes is to obtain a global model. It should be able to capture the basic or general behavior of the system, by means of a linear or quadratic regression, and then superimpose a local model on it that can capture the localized nonlinearities of the system. In this paper, a novel method based on a hybrid incremental modeling approach is designed and applied for tool wear detection in turning processes. It involves a two-step iterative process that combines a global model with a local model to take advantage of their underlying, complementary capacities. Thus, the first step constructs a global model using a least squares regression. A local model using the fuzzy k-nearest-neighbors smoothing algorithm is obtained in the second step. A comparative study then demonstrates that the hybrid incremental model provides better error-based performance indices for detecting tool wear than a transductive neurofuzzy model and an inductive neurofuzzy model.

Content Based Video Retrieval using SURF Descriptor

This paper presents a Robust Content Based Video Retrieval (CBVR) system. This system retrieves similar videos based on a local feature descriptor called SURF (Speeded Up Robust Feature). The higher dimensionality of SURF like feature descriptors causes huge storage consumption during indexing of video information. To achieve a dimensionality reduction on the SURF feature descriptor, this system employs a stochastic dimensionality reduction method and thus provides a model data for the videos. On retrieval, the model data of the test clip is classified to its similar videos using a minimum distance classifier. The performance of this system is evaluated using two different minimum distance classifiers during the retrieval stage. The experimental analyses performed on the system shows that the system has a retrieval performance of 78%. This system also analyses the performance efficiency of the low dimensional SURF descriptor.

Evaluation of Speaker Normalization Methods for Vowel Recognition Using Fuzzy ARTMAP and K-NN

A procedure that uses fuzzy ARTMAP and K-Nearest Neighbor (K-NN) categorizers to evaluate intrinsic and extrinsic speaker normalization methods is described. Each classifier is trained on preprocessed, or normalized, vowel tokens from about 30% of the speakers of the Peterson-Barney database, then tested on data from the remaining speakers. Intrinsic normalization methods included one nonscaled, four psychophysical scales (bark, bark with end-correction, mel, ERB), and three log scales, each tested on four different combinations of the fundamental (Fo) and the formants (F1 , F2, F3). For each scale and frequency combination, four extrinsic speaker adaptation schemes were tested: centroid subtraction across all frequencies (CS), centroid subtraction for each frequency (CSi), linear scale (LS), and linear transformation (LT). A total of 32 intrinsic and 128 extrinsic methods were thus compared. Fuzzy ARTMAP and K-NN showed similar trends, with K-NN performing somewhat better and fuzzy ARTMAP requiring about 1/10 as much memory. The optimal intrinsic normalization method was bark scale, or bark with end-correction, using the differences between all frequencies (Diff All). The order of performance for the extrinsic methods was LT, CSi, LS, and CS, with fuzzy AHTMAP performing best using bark scale with Diff All; and K-NN choosing psychophysical measures for all except CSi.

The condensed nearest neighbor rule using the concept of mutual nearest neighborhood (Corresp.)

A two-stage iterative algorithm for selecting a subset of a training set of samples for use in a condensed nearest neighbor (CNN) decision rule is introduced. The proposed method uses the concept of mutual nearest neighborhood for selecting samples close to the decision line. The efficacy of the algorithm is brought out by means of an example.

Criticality of the exponential rate of decay for the largest nearest-neighbor link in random geometric graphs

Let n points be placed independently in d-dimensional space according to the density f(x) = A(d)e(-lambda parallel to x parallel to alpha), lambda, alpha > 0, x is an element of R-d, d >= 2. Let d(n) be the longest edge length of the nearest-neighbor graph on these points. We show that (lambda(-1) log n)(1-1/alpha) d(n) - b(n) converges weakly to the Gumbel distribution, where b(n) similar to ((d - 1)/lambda alpha) log log n. We also prove the following strong law for the normalized nearest-neighbor distance (d) over tilde (n) = (lambda(-1) log n)(1-1/alpha) d(n)/log log n: (d - 1)/alpha lambda <= lim inf(n ->infinity) (d) over tilde (n) <= lim sup(n ->infinity) (d) over tilde (n) <= d/alpha lambda almost surely. Thus, the exponential rate of decay alpha = 1 is critical, in the sense that, for alpha > 1, d(n) -> 0, whereas, for alpha <= 1, d(n) -> infinity almost surely as n -> infinity.