No description available.

The computer classification of musical audio can form the basis for systems that allow new ways of interacting with digital music collections. Existing music classification systems suffer, however, from inaccuracy as well as poor scalability. Feature selection is a machine-learning tool that can potentially improve both accuracy and scalability of classification. Unfortunately, there is no consensus on which feature selection algorithms are most appropriate or on how to evaluate the effectiveness of feature selection. Based on relevant literature in music information retrieval (MIR) and machine learning and on empirical testing, the thesis specifies an appropriate evaluation method for feature selection, employs this method to compare existing feature selection algorithms, and evaluates an appropriate feature selection algorithm on the problem of musical genre classification. The outcomes include an increased understanding of the potential for feature selection to benefit MIR and a new technique for optimizing one type of classification-based system.

The goal of this research was to explore the use of machine learning to assist in the development of knowledge-based systems (KBS) in dairy farming. A framework was first developed which described the various types of management and control activities in dairy farming and the types of information flows among these activities. This framework provided a basis for the creation of computerized information systems and helped to identify the analysis of group-average lactation curves as a promising area of application. A case-acquisition and decision-support system was developed to assist a domain specialist in generating example cases for machine learning. The specialist classified data from 33 herds enrolled with the Quebec dairy herd analysis service, resulting in 1428 lactations and 7684 tests of individual cows, classified as outlier or non-outlier, and 99 interpretations of group-average lactation curves. To enable the performance analysis of classifiers, generated with machine learning from these small data sets, a method was established involving cross-validation runs, relative operating characteristic curves, and analysis of variance. In experiments to filter lactations and tests, classification performance was significantly affected by preprocessing of examples, creation of additional attributes, choice of machine-learning algorithm, and algorithm configuration. For the filtering of individual tests, naive-Bayes classification showed significantly better performance than decision-tree induction. However, the specialist considered the decision trees as more transparent than the knowledge generated with naive Bayes. The creation of a series of three classifiers with increased sensitivity at the expense of reduced specificity per classification task, allows users of a final KBS to choose the desired tendency of classifying new cases as abnormal. For the main interpretation tasks, satisfactory performance was achieved. For the filtering tasks, performance was fai

<p> <p>&nbsp / </p> </p> <p align="left">IP multicast is an efficient mechanism for simultaneously transmitting bulk data to multiple receivers. Many applications can benefit from multicast, such as audio and videoconferencing, multi-player games, multimedia broadcasting, distance education, and data replication. For either technical or policy reasons, IP multicast still has not yet been deployed in today&rsquo / s Internet. Congestion is one of the most important issues impeding the development and deployment of IP multicast and multicast applications.</p>

New methods of attacking networks are being invented at an alarming rate, and pure signature detection cannot keep up. The ability of intrusion detection systems to generalize to new attacks based on behavior is of increasing value. Machine Learning algorithms have been successfully applied to intrusion and fraud detection; however the time and accuracy tradeoffs between algorithms are not always considered when faced with such a broad range of choices. This thesis explores the time and accuracy metrics of a wide variety of machine learning algorithms, using a purpose-built supervised learning dataset. Topics covered include dataset dimensionality reduction through pre-processing techniques, training and testing times, classification accuracy, and performance tradeoffs. Further, ensemble learning and meta-classification are used to explore combinations of the algorithms and derived data sets, to examine the effects of homogeneous and heterogeneous aggregations. The results of this research are presented with observations and guidelines for choosing learning schemes in this domain.

Cancer is a complex and deadly disease that is caused by genetic lesions in somatic cells. Further research in computational methodology for detecting and characterizing somatic mutations is necessary in order to understand the comprehensive systems level model of the roles of those lesions in cancer development. In the first project, I trained a list of supervised machine learning classifiers that classify false positive versus true positive somatic single nucleotide variants (SNVs). I was able to show an improvement of somatic SNV detection on the data set over the reported classifier. In the second project, we developed PhyloSub model that uses a nonparametric Bayesian prior over a set of trees to cluster SNVs, and infer the subclonal phylogenetic structure of tumors with uncertainty from SNV sequencing data. Experiments showed that PhyloSub model could infer the subclonal phylogenetic structure from both single and multiple tumor samples.

Cancer is a complex and deadly disease that is caused by genetic lesions in somatic cells. Further research in computational methodology for detecting and characterizing somatic mutations is necessary in order to understand the comprehensive systems level model of the roles of those lesions in cancer development. In the first project, I trained a list of supervised machine learning classifiers that classify false positive versus true positive somatic single nucleotide variants (SNVs). I was able to show an improvement of somatic SNV detection on the data set over the reported classifier. In the second project, we developed PhyloSub model that uses a nonparametric Bayesian prior over a set of trees to cluster SNVs, and infer the subclonal phylogenetic structure of tumors with uncertainty from SNV sequencing data. Experiments showed that PhyloSub model could infer the subclonal phylogenetic structure from both single and multiple tumor samples.

We explore both robust biologically guided intensity-modulated radiation therapy (BG-IMRT) and pattern recognition to identify responders to cancer treatment for lung cancer. Heterogeneous dose prescriptions that are derived from biological images are subject to uncertainty, due to potential noise in the image. We develop a robust optimization model to design BG-IMRT plans that are de-sensitized to uncertainty. Computational results show improvements in tumor control probability and deviation from prescription dose compared to a non-robust model, while maintaining tissue dose below toxicity levels. We applied machine learning algorithms to 4D gated positron emission tomography/computed tomography (PET/CT) scans. We identified classifiers which could outperform a naive classifier. Our work shows the potential of using machine learning algorithms to predict patient response. This could hopefully lead to more adaptive treatment plans, where the clinician would adapt the treatment based on the prediction provided at certain time intervals in the treatment.

We explore both robust biologically guided intensity-modulated radiation therapy (BG-IMRT) and pattern recognition to identify responders to cancer treatment for lung cancer. Heterogeneous dose prescriptions that are derived from biological images are subject to uncertainty, due to potential noise in the image. We develop a robust optimization model to design BG-IMRT plans that are de-sensitized to uncertainty. Computational results show improvements in tumor control probability and deviation from prescription dose compared to a non-robust model, while maintaining tissue dose below toxicity levels. We applied machine learning algorithms to 4D gated positron emission tomography/computed tomography (PET/CT) scans. We identified classifiers which could outperform a naive classifier. Our work shows the potential of using machine learning algorithms to predict patient response. This could hopefully lead to more adaptive treatment plans, where the clinician would adapt the treatment based on the prediction provided at certain time intervals in the treatment.

No description available.