Fusion of multiple approximate nearest neighbor classifiers for fast and efficient classification

2004 ◽  
Vol 5 (4) ◽  
pp. 239-250 ◽  
Author(s):  
P. Viswanath ◽  
M. Narasimha Murty ◽  
Shalabh Bhatnagar
Keyword(s):  
Nearest Neighbor ◽  
Approximate Nearest Neighbor ◽  
Nearest Neighbor Classifiers

scholarly journals Semantics of Voids within Data: Ignorance-Aware Machine Learning

2021 ◽  
Vol 10 (4) ◽  
pp. 246
Author(s):  
Vagan Terziyan ◽  
Anton Nikulin
Keyword(s):  
Spatial Data ◽  
Information Science ◽  
Nearest Neighbor ◽  
Geographical Information ◽  
Two Dimensional ◽  
Prototype Selection ◽  
Data Space ◽  
Important Concern ◽  
Traditional Classification ◽  
Nearest Neighbor Classifiers

Operating with ignorance is an important concern of geographical information science when the objective is to discover knowledge from the imperfect spatial data. Data mining (driven by knowledge discovery tools) is about processing available (observed, known, and understood) samples of data aiming to build a model (e.g., a classifier) to handle data samples that are not yet observed, known, or understood. These tools traditionally take semantically labeled samples of the available data (known facts) as an input for learning. We want to challenge the indispensability of this approach, and we suggest considering the things the other way around. What if the task would be as follows: how to build a model based on the semantics of our ignorance, i.e., by processing the shape of “voids” within the available data space? Can we improve traditional classification by also modeling the ignorance? In this paper, we provide some algorithms for the discovery and visualization of the ignorance zones in two-dimensional data spaces and design two ignorance-aware smart prototype selection techniques (incremental and adversarial) to improve the performance of the nearest neighbor classifiers. We present experiments with artificial and real datasets to test the concept of the usefulness of ignorance semantics discovery.

Large-margin nearest neighbor classifiers via sample weight learning

2011 ◽  
Vol 74 (4) ◽  
pp. 656-660 ◽  
Author(s):  
Qinghua Hu ◽  
Pengfei Zhu ◽  
Yongbin Yang ◽  
Daren Yu
Keyword(s):  
Nearest Neighbor ◽  
Sample Weight ◽  
Large Margin ◽  
Weight Learning ◽  
Nearest Neighbor Classifiers

scholarly journals Pengenalan Motif Karawo Menggunakan Ekstraksi Fitur SIFT Dan Approximate Nearest Neighbor

2019 ◽  
Author(s):  
Syahrial
Keyword(s):  
Nearest Neighbor ◽  
Feature Matching ◽  
Recognition Accuracy ◽  
Sift Algorithm ◽  
Approximate Nearest Neighbor ◽  
Single Pattern ◽  
Tree Data ◽  
Threshold Testing ◽  
Root Word ◽  
Tree Data Structure

An art culture from Gorontalo became iconic handcraft is kerawang or karawo. The word “karawo” came from root word of “mokarawo” which means slicing or making holes. It’s created with precision, carefulness, and patience in work using handmade masterpiece. Pattern of karawo itself held four kinds which is flora, fauna, geometric, and nature. From those kinds born vary pattern which come difficult to identify both its names and its kind. Karawo patterns can be form as a single pattern or a pattern that it parts came from several or many pattern combined. Those patterns had its own characteristic from shape and scale perspective. Identifying single pattern on combined pattern are particularly a problem because it’s combined involve scaling and rotation. This research is recognizing single pattern on combined pattern using feature extraction SIFT algorithm which is capable extract feature that invariant from scale and rotation. Feature matching using approximate nearest neighbor (aNN) for similarity of features labor best bin first strategy on kd-tree data structure. Those methods can be a reference to recognize single pattern on combined pattern using from range 5 to 20 match features as a threshold. Testing result indicated recognition accuracy is good which range form 76.36% to 85.45% on recognize the kind of karawo pattern and 76.36% on its name.

Sign in / Sign up

Export Citation Format

Share Document