Multi-task learning improves generalization performance by learning several related tasks jointly. Several methods have been proposed for multi-task learning in recent years. Many methods make strong assumptions about symmetric task relationships while some are able to utilize externally provided task relationships. However, in many real world tasks the degree of relatedness among tasks is not known a priori. Methods which are able to extract the task relationships and exploit them while simultaneously learning models with good generalization performance can address this limitation. In the current work, we have extended a recently proposed method for learning task relationships using smooth squared loss for regression to classification problems using non-smooth hinge loss due to the demonstrated effectiveness of SVM classifier in single task classification settings. We have also developed an efficient optimization procedure using bundle methods for the proposed multi-task learning formulation. We have validated our method on one simulated and two real world datasets and have compared its performance to competitive baseline single-task and multi-task methods.

Classification problems in which several learning tasks are organized hierarchically pose a special challenge because the hierarchical structure of the problems needs to be considered. Multi-task learning (MTL) provides a framework for dealing with such interrelated learning tasks. When two different hierarchical sources organize similar information, in principle, this combined knowledge can be exploited to further improve classification performance. We have studied this problem in the context of protein structure classification by integrating the learning process for two hierarchical protein structure classification database, SCOP and CATH. Our goal is to accurately predict whether a given protein belongs to a particular class in these hierarchies using only the amino acid sequences. We have utilized the recent developments in multi-task learning to solve the interrelated classification problems. We have also evaluated how the various relationships between tasks affect the classification performance. Our evaluations show that learning schemes in which both the classification databases are used outperform the schemes which utilize only one of them.

Multi-task learning (MTL) is a supervised learning paradigm in which the prediction models for several related tasks are learned jointly to achieve better generalization performance. When there are only a few training examples per task, MTL considerably outperforms the traditional Single task learning (STL) in terms of prediction accuracy. In this work we develop an MTL based approach for classifying documents that are archived within dual concept hierarchies, namely, DMOZ and Wikipedia. We solve the multi-class classification problem by defining one-versus-rest binary classification tasks for each of the different classes across the two hierarchical datasets. Instead of learning a linear discriminant for each of the different tasks independently, we use a MTL approach with relationships between the different tasks across the datasets established using the non-parametric, lazy, nearest neighbor approach. We also develop and evaluate a transfer learning (TL) approach and compare the MTL (and TL) methods against the standard single task learning and semi-supervised learning approaches. Our empirical results demonstrate the strength of our developed methods that show an improvement especially when there are fewer number of training examples per classification task.

Several biological databases organize information in taxonomies/hierarchies. These databases differ in terms of curation process, input data, coverage and annotation errors. SCOP and CATH are examples of two databases that classify proteins hierarchically into structurally related groups based on experimentally determined structures. Given the large number of protein sequences with unavailable structure, there is a need to develop prediction methods to classify protein sequences into structural classes. We have developed a novel classification approach that utilizes the underlying relationships across multiple hierarchical source databases within a multi-task learning (MTL) framework. MTL is used to simultaneously learn multiple related tasks, and has been shown to improve generalization performance. Specifically, we have developed and evaluated an MTL approach for predicting the structural class, as defined by two hierarchical databases, CATH and SCOP, using protein sequence information only. We define one task per node of the hierarchies and formulate the MTL problem as a combination of these binary classification tasks. Our experimental evaluation demonstrates that the MTL approach that integrates both the hierarchies outperforms the base-line approach that trains independent models per task, as well as a MTL approach that integrates tasks across a single hierarchical database. We also performed extensive experiments that evaluate different regularization penalties and incorporate different task relationships that achieve superior classification performance.

Assembling short reads obtained from community samples using next-generation sequencing technologies is challenging due to several reasons. In this study we assess the performance of a state-of-the-art Eulerian-path based assembler on a series of simulated dataset with varying complexities. We evaluate the feasibility of metagenomic assembly with reads restricted to length 36 base pairs, obtained from the Solexa/Illumina platform. We developed a pipeline to evaluate the quality of assembly based on contig length statistics and accuracy. We studied the effect of overlap parameters used for the metagenomic assembly and developed a clustering solution to pool the contigs obtained from different runs of the assembly algorithm which allowed us to obtain longer contigs from different runs. We also computed an entropy/impurity metric to assess the degree of chimericity in the assembled contigs. We also compared the metagenomic assemblies to the best possible solution that could be obtained by assembling individual source genomes. Our results show that accuracy was better than expected for the metagenomic samples with a few dominant organisms and was especially poor in samples containing many closely related strains.