Traditionally, supervised machine learning methods are the first choice for tasks involving classification of data. This study provides a non-conventional hybrid alternative technique (pEAC) that blends the Possibilistic Fuzzy CMeans (PFCM) as base cluster generating algorithm into the ‘standard’ Evidence Accumulation Clustering (EAC) clustering method. The PFCM coalesces the separate propertie...

In land use planning, mapping the present land use / land cover situation is a necessary tool for determining the current condition and for identifying land use trends. In this study, in order to provide a land use/ land cover map for Ilam watershed, the IRS-1C image data from 25th April 2006 were used. Initial qualitative evaluation on data showed no significant radiometric error. Ortho-rectif...

Image segmentation is the process of finding out all non-overlapping distinct regions from the given image based on certain criteria such as intensity, color, texture or shape. This paper proposes a two level hybrid non classical model for image segmentation based on pixel color and texture features of the image. The first level uses Fuzzy C-Means (FCM) unsupervised method to form a clustering ...

This paper presents the design of a system for feature extraction and classification of news articles from Croatian news sources. An overview of supervised and unsupervised text classification and clustering machine learning techniques is presented. The techniques described are those most widely used for text classification tasks. The paper discusses a number of issues particular to text classi...

Unsupervised learning and supervised learning are key research topics in deep learning. However, as high-capacity supervised neural networks trained with a large amount of labels have achieved remarkable success in many computer vision tasks, the availability of large-scale labeled images reduced the significance of unsupervised learning. Inspired by the recent trend toward revisiting the impor...

Ensemble classifiers such as bagging, boosting and model averaging are known to have improved accuracy and robustness over a single model. Their potential, however, is limited in applications which have no access to raw data but to the meta-level model output. In this paper, we study ensemble learning with output from multiple supervised and unsupervised models, a topic where little work has be...

This paper addresses the problem of unsupervised object localization in an image. Unlike previous supervised and weakly supervised algorithms that require bounding box or image level annotations for training classifiers in order to learn features representing the object, we propose a simple yet effective technique for localization using iterative spectral clustering. This iterative spectral clu...

We present a novel unsupervised approach for multilingual sentiment analysis driven by compositional syntax-based rules. On the one hand, we exploit some of the main advantages of unsupervised algorithms: (1) the interpretability of their output, in contrast with most supervised models, which behave as a black box and (2) their robustness across different corpora and domains. On the other hand,...

This paper deals with the supervised classification when the response variable is binary and its class distribution is unbalanced. In such situation, it is not possible to build a powerful classifier by using standard methods such as logistic regression, classification tree, discriminant analysis, etc. To overcome this shortcoming of these methods that provide classifiers with low sensibility, ...

Modification of the original bands and integration of ancillary data in digital image classification has been shown to improve land use land cover classification accuracy. There are not many studies demonstrating such techniques in the context of the mountains of Nepal. The objective of this study was to explore and evaluate the use of modified band and ancillary data in Landsat and IRS image c...