While k-nearest neighbor queries are becoming increasingly common due to mobile and geospatial applications, orthogonal range queries in high-dimensional data are extremely important in scientific and web-based applications. For efficient querying, data is typically stored in an index optimized for either kNN or range queries. This can be problematic when data is optimized for kNN retrieval and...

This paper describes a machine learning method, called Regression by Feature Projections (RFP), for predicting a real-valued target feature. In RFP training is based on simply storing the projections of the training instances on each feature separately. Prediction is computed through two approximation procedures. The first approximation process is to find the individual predictions of features ...

In this paper, a novel prototype reduction algorithm is proposed, which aims at reducing the storage requirement and enhancing the online speed while retaining the same level of accuracy for a K-nearest neighbor (KNN) classifier. To achieve this goal, our proposed algorithm learns the weighted similarity function for a KNN classifier by maximizing the leave-one-out cross-validation accuracy. Un...

Indoor Positioning System aims at locating objects inside buildings wirelessly, and have huge benefit for indoor location-aware mobile application. To explore this immature system design, we choose UJIndoorLoc database as our data set, use PCA for feature selection, and build prediction models based on decision tree, gradient boosting, kNN and SVM, respectively. Our experiment results indicate ...

Multi-Feature Index Tree (MFI-Tree), a new indexing structure, is proposed to index multiple high-dimensional features of video data for video retrieval through example. MFI-Tree employs tree structure which is beneficial for the browsing application, and retrieves the last level cluster nodes in retrieval application to improve the performance. Aggressive Decided Distance for kNN (ADD-kNN) sea...

Die Anforderungen an die moderne Landwirtschaft bedingen einen effizienten Ressourceneinsatz, auch bei der N-Düngung. Ökonomisch optimierte Entscheidungsregeln wurden mit Künstlichen Neuronalen Netzen (KNN) und Support-Vector-Machines (SVM) erzeugt und in der Anwendung mit einer betriebseinheitlichen Variante verglichen. Es zeigte sich, dass Einsparpotentiale von bis zu 69 kg N/ha möglich sind ...

Given a query point q and a set D of data points, a nearest neighbor (NN) query returns the data point p in D that minimizes the distance DIST(q,p), where the distance function DIST(,) is the L2 norm. One important variant of this query type is kNN query, which returns k data points with the minimum distances. When taking the temporal dimension into account, the kNN query result may change over...

K Nearest Neighbor (kNN) is one of the most popular machine learning techniques, but it often fails to work well with inappropriate choice of distance metric or due to the presence of a lot of irrelated features. Linear and non-linear feature transformation methods have been applied to extract classrelevant information to improve kNN classification. In this paper, I describe kNN classification ...

K-nearest neighbor graph (KNN) is a widely used tool in several pattern recognition applications but it has drawbacks. Firstly, the choice of k can have significant impact on the result because it has to be fixed beforehand, and it does not adapt to the local density of the neighborhood. Secondly, KNN does not guarantee connectivity of the graph. We introduce an alternative data structure calle...

The Internet provides easy access to a kind of library resources. However, classification of documents from a large amount of data is still an issue and demands time and energy to find certain documents. Classification of similar documents in specific classes of data can reduce the time for searching the required data, particularly text documents. This is further facilitated by using Artificial...