Improving Performance of Self-Organising Maps with Distance Metric Learning Method

نویسندگان

  • Piotr Plonski
  • Krzysztof Zaremba
چکیده

Self-Organising Maps (SOM) are Artificial Neural Networks used in Pattern Recognition tasks. Their major advantage over other architectures is human readability of a model. However, they often gain poorer accuracy. Mostly used metric in SOM is the Euclidean distance, which is not the best approach to some problems. In this paper, we study an impact of the metric change on the SOM’s performance in classification problems. In order to change the metric of the SOM we applied a distance metric learning method, so-called ’Large Margin Nearest Neighbour’. It computes the Mahalanobis matrix, which assures small distance between nearest neighbour points from the same class and separation of points belonging to different classes by large margin. Results are presented on several real data sets, containing for example recognition of written digits, spoken letters or faces.

برای دانلود متن کامل این مقاله و بیش از 32 میلیون مقاله دیگر ابتدا ثبت نام کنید

ثبت نام

اگر عضو سایت هستید لطفا وارد حساب کاربری خود شوید

منابع مشابه

Composite Kernel Optimization in Semi-Supervised Metric

Machine-learning solutions to classification, clustering and matching problems critically depend on the adopted metric, which in the past was selected heuristically. In the last decade, it has been demonstrated that an appropriate metric can be learnt from data, resulting in superior performance as compared with traditional metrics. This has recently stimulated a considerable interest in the to...

متن کامل

A Ciric-type common fixed point theorem in complete b-metric spaces

In this paper, we define the concept of compatible and weakly compatible mappings in b-metric spaces and inspired by the Ciric and et. al method, we produce appropriate conditions for given the unique common fixed point for a family of the even number of self-maps with together another two self-maps in a complete b-metric spaces. Also, we generalize this common fixed point theorem for a seq...

متن کامل

Comparative Impact of In-Presence and Distance Learning Methods on Knowledge Improvement in the Healthcare Workers

Background and Objectives: In continuous medical learning programs, selection of appropriate learning methods is important for streaming individuals by learning process. The goal of this study is to compare in-presence learning method with distance learning method and explore their effect on the knowledge of health workers (practical nurses) in terms of reduced expenses.   Methods: The study ...

متن کامل

یادگیری نیمه نظارتی کرنل مرکب با استفاده از تکنیک‌های یادگیری معیار فاصله

Distance metric has a key role in many machine learning and computer vision algorithms so that choosing an appropriate distance metric has a direct effect on the performance of such algorithms. Recently, distance metric learning using labeled data or other available supervisory information has become a very active research area in machine learning applications. Studies in this area have shown t...

متن کامل

Middle Scale Robot Navigation -a Case Study

In this paper we present results of experiments carried out with a route learning system for a mobile robot, conducted in a `real world' environment covering distances of several hundred metres. The system uses no odometry and is based on a self-organising mapbuilding process using perceptual landmarks. A performance metric is deened and used to measure the robot's ability to traverse the route.

متن کامل

ذخیره در منابع من


  با ذخیره ی این منبع در منابع من، دسترسی به آن را برای استفاده های بعدی آسان تر کنید

برای دانلود متن کامل این مقاله و بیش از 32 میلیون مقاله دیگر ابتدا ثبت نام کنید

ثبت نام

اگر عضو سایت هستید لطفا وارد حساب کاربری خود شوید

عنوان ژورنال:

دوره   شماره 

صفحات  -

تاریخ انتشار 2012