Automatic detection of craters in planetary images an embedded framework using feature selection and boosting.


Learning and Meta%2DLearning for Coordination of Autonomous Unmanned Vehicles %2D A Preliminary Analysis.


Meta%2DLearning %2D Concepts and Techniques.

A unified framework for reinforcement learning

A Framework for Multi%2DObjective Clustering and Its Application to Co%2DLocation Mining.
Abstract. The goal of multi-objective clustering (MOC) is to decompose a dataset into similar groups maximizing multiple objectives in parallel. In this paper, we provide a methodology, architecture and algorithms that, based on a large set of objectives, derive interesting clusters regarding two or more of those objectives. The proposed architecture relies on clustering algorithms that support plug-in fitness functions and on multi-run clustering in which clustering algorithms are run multiple times maximizing different subsets of objectives that are captured in compound fitness functions. MOC provides search engine type capabilities to users, enabling them to query a large set of clusters with respect to different objectives and thresholds. We evaluate the proposed MOC framework in a case study that centers on spatial co-location mining; the goal is to identify regions in which high levels of Arsenic concentrations are co-located with high concentrations of other chemicals in the Texas water supply. Keywords: Multi-objective clustering, knowledge discovery, spatial data mining, co-location mining, clustering with plug-in fitness functions. 1


Data Selection Using SASH Trees for Support Vector Machines.

Tablet PC video based hybrid coursework in computer science report from a pilot project.
Online learning, defined broadly as recording and delivering classroom experience with technology, has tremendous potential. However, success to date has been very limited in science and engineering. We believe this is because traditional video recording is cumbersome and not suitable for technical lectures and removing live classroom interaction is detrimental to learning. Employing Tablet PCs with slide presentation software has made it simple and convenient to develop and record high quality lectures. We employ such videos in a hybrid model of coursework. All lectures are made available as online videos, but limited classroom interaction is an important component; the classroom is used for review of lecture material, examinations, demonstrations, tutorials, and hands-on sessions. The hybrid framework is particularly suitable for students with logistical difficulties, e.g., because of work schedule. This paper is an evaluation of the hybrid learning approach as applied to upper level computer science coursework. We report our experience in teaching a suite of hybrid courses at the University of Houston and discuss the detailed feedback we received from the students who participated in the courses.
An efficient approach to external cluster assessment with an application to martian topography.

Machine Learning Tools for Automatic Mapping of Martian Landforms.

Automatic Recognition of Landforms on Mars Using Terrain Segmentation and Classification.
Abstract. Mars probes send back to Earth enormous amount of data. Automating the analysis of this data and its interpretation represents a challenging test of significant benefit to the domain of planetary science. In this study, we propose combining terrain segmentation and classification to interpret Martian topography data and to identify constituent landforms of the Martian landscape. Our approach uses unsupervised segmentation to divide a landscape into a number of spatially extended but topographically homogeneous objects. Each object is assigned a 12 dimensional feature vector consisting of terrain attributes and neighborhood properties. The objects are classified, based on their feature vectors, into predetermined landform classes. We have applied our technique to the Tisia Valles test site on Mars. Support Vector Machines produced the most accurate results (84.6 % mean accuracy) in the classification of topographic objects. An immediate application of our algorithm lies in the automatic detection and characterization of craters on Mars. 1
Using clustering to learn distance functions for supervised similarity assessment.

Adaptive Clustering Obtaining Better Clusters Using Feedback and Past Experience.
Adaptive clustering uses external feedback to improve cluster quality; past experience serves to speed up execution time. An adaptive clustering environment is proposed that uses Q-learning to learn the reward values of successive data clusterings. Adaptive clustering supports the reuse of clusterings by memorizing what worked well in the past. It has the capability of exploring multiple paths in parallel when searching for good clusters. In a case study, we apply adaptive clustering to instance-based learning relying on a distance function modification approach. A distance function adaptation scheme that uses external feedback is proposed and compared with other distance function learning approaches. Experimental results indicate that the use of adaptive clustering leads to significant improvements of instance-based learning techniques, such as k-nearest neighbor classifiers. Moreover, as a by-product a new instance-based learning technique is introduced that classifies examples by solely using cluster representatives; this technique shows high promise in our experimental evaluation. 1.
Content%2DBased Image Retrieval through a Multi%2DAgent Meta%2DLearning Framework.
The objective of a general-purpose content-based image retrieval system is to find images in a database that match an external measure of relevance. Since users follow different and inconsistent relevance measures, processing queries in a task-specific manner has shown to be an effective approach. Viewing specialized image retrieval algorithms as agents, we propose a general-purpose image retrieval system that uses a new multi-agent meta-learning framework. The framework adapts a distance function defined over both image distance weights and image queries to identify clusters of algorithms that produce similar solutions to similar problems. Experiments compare our approach with a traditional information retrieval algorithm; results show that our framework provides better average relevance scores. 1.
Using Clustering to Learn Distance Functions for Supervised Similarity Assessment.

Testing Theories in Particle Physics Using Maximum Likelihood and Adaptive Bin Allocation.

Meta%2DLearning.
as scheme-based search with complexity control
Piece%2DWise Model Fitting Using Local Data Patterns.
Abstract. In this paper we propose a novel classification algorithm that fits models of different complexity on separate regions of the input space. The goal is to achieve a balance between global and local learning strategies by decomposing the classification task into simpler subproblems; each task narrows the learning problem to a local region of high example density over the input space. Specifically, our proposed approach is to apply a clustering algorithm to every set of training examples that belong to the same class; each cluster becomes an intermediate concept that is learned by selecting a model with an (estimated) optimal degree of complexity. Experimental results on real-world domains show consistent good performance in predictive accuracy with our piece-wise model fitting strategy. 1
Using Representative%2DBased Clustering for Nearest Neighbor Dataset Editing.
The goal of dataset editing in instance-based learning is to remove objects from a training set in order to increase the accuracy of a classifier. For example, Wilson editing removes training examples that are misclassified by a nearest neighbor classifier so as to smooth the shape of the resulting decision boundaries. This paper revolves around the use of representative-based clustering algorithms for nearest neighbor dataset editing. We term this approach supervised clustering editing. The main idea is to replace a dataset by a set of cluster prototypes. A novel clustering approach called supervised clustering is introduced for this purpose. Our empirical evaluation using eight UCI datasets shows that both Wilson and supervised clustering editing improve accuracy on more than 50 % of the datasets tested. However, supervised clustering editing achieves four times higher compression rates than Wilson editing; moreover, it obtains significantly high accuracies for three of the eight datasets tested.
A Quantification of Cluster Novelty with an Application to Martian Topography.



Introduction to the Special Issue on Meta%2DLearning.

A Decomposition of Classes via Clustering to Explain and Improve Naive Bayes.
Abstract. We propose a method to improve the probability estimates made by Naive Bayes to avoid the effects of poor class conditional probabilities based on product distributions when each class spreads into multiple regions. Our approach is based on applying a clustering algorithm to each subset of examples that belong to the same class, and to consider each cluster as a class of its own. Experiments on 26 real-world datasets show a significant improvement in performance when the class decomposition process is applied, particularly when the mean number of clusters per class is large. 1
Class Decomposition via Clustering A New Framework for Low%2DVariance Classifiers.
In this paper we propose a pre-processing step to classification that applies a clustering algorithm to the training set to discover local patterns in the attribute or input space. We demonstrate how this knowledge can be exploited to enhance the predictive accuracy of simple classifiers. Our focus is mainly on classifiers characterized by high bias but low variance (e.g., linear classifiers); these classifiers experience difficulty in delineating class boundaries over the input space when a class distributes in complex ways. Decomposing classes into clusters makes the new class distribution easier to approximate and provides a viable way to reduce bias while limiting the growth in variance. Experimental results on real-world domains show an advantage in predictive accuracy when clustering is used as a preprocessing step to classification. 1
Critical event prediction for proactive management in large%2Dscale computer clusters.

Evaluation Metrics in Classification A Quantification of Distance%2DBias.

Predicting Rare Events In Temporal Domains.

A Characterization of Difficult Problems in Classification.
Most classification algorithms experience difficulties when the input-output distribution is irregular (i.e., neighbor examples in the input space have different values in the output space), and sparse. We characterize the difficulty of a classification problem using a limited number of examples by re-defining a measure of class variation to account for irregular distributions, and by introducing a new measure of example cohesiveness in the input space to account for (lack of) sparseness. Our empirical study indicates these measures correlate with accuracy performance.
A Classification Approach for Prediction of Target Events in Temporal Sequences.
Learning to predict significant events from sequences of data  with categorical features is an important problem in many application areas.
A Perspective View and Survey of Meta%2DLearning.
Abstract. Different researchers hold different views of what the term meta-learning exactly means. The first part of this paper provides our own perspective view in which the goal is to build self-adaptive learners (i.e. learning algorithms that improve their bias dynamically through experience by accumulating meta-knowledge). The second part provides a survey of meta-learning as reported by the machine-learning literature. We find that, despite different views and research lines, a question remains constant: how can we exploit knowledge about learning (i.e. meta-knowledge) to improve the performance of learning algorithms? Clearly the answer to this question is key to the advancement of the field and continues being the subject of intensive research.
Predictive algorithms in the management of computer systems.

A Unified Framework for Evaluation Metrics in Classification Using Decision Trees.
. Most evaluation metrics in classication are designed to reward  class uniformity in the example subsets induced by a feature (e.g.,  Information Gain). Other metrics are designed to reward discrimination  power in the context of feature selection as a means to combat the  feature-interaction problem (e.g., Relief, Contextual Merit). We dene  a new framework that combines the strengths of both kinds of metrics.  Our framework enriches the available information when considering  which feature to use to partition the training set. Since most metrics  rely on only a small fraction of this information, this framework enlarges  the space of possible metrics. Experiments on real-world domains in the  context of decision-tree learning show how a simple setting for our framework  compares well with standard metrics.  1 
Operational Data Analysis Improved Predictions Using Multi%2Dcomputer Pattern Detection.
Operational Data Analysis (ODA) automatically 1) monitors the performance of a computer through time, 2) stores such information in a data repository, 3) applies data-mining techniques, and 4) generates results. We describe a system implementing the four steps in ODA, focusing our attention on the data-mining step where our goal is to predict the value of a performance parameter (e.g., response time, cpu utilization, memory utilization) in the future.
A Quantification of Distance Bias Between Evaluation Metrics In Classification.
This paper provides a characterization of bias  for evaluation metrics in classification (e.g.,  Information Gain, Gini,   2  , etc.). Our characterization  provides a uniform representation  for all traditional evaluation metrics.  Such representation leads naturally to a measure  for the distance between the bias of  two evaluation metrics. We give a practical  value to our measure by observing if  the distance between the bias of two evaluation  metrics correlates with differences in  predictive accuracy when we compare two  versions of the same learning algorithm that  differ in the evaluation metric only. Experiments  on real-world domains show how  the expectations on accuracy differences generated  by the distance-bias measure correlate  with actual differences when the learning  algorithm is simple (e.g., search for the  best single-feature or the best single-rule).  The correlation, however, weakens with more  complex algorithms (e.g., learning decision  trees). Our results sh...


