|
advertisement
|
|
|
| Book Lists Home - Bioinformatics Books - Add To My Book List |
|
|
Book Summary:
The completion of the sequencing of the human genome was heralded the dawn of a new era in biology and medicine. Besides, advances in microarray technologies enable us to simultaneously observe the expression levels of many thousands of genes on the transcription levels during important biological processes. Such a global view of thousands of functional genes also changes the landscape of biological and biomedical research. However, the huge amount of DNA sequence data and gene expression data are of limited value if we cannot use them to discover the function and regulation of gene products. This thesis is devoted to developing effective and efficient machine learning techniques for analyzing the huge amount of genetic data. Mathematically, learning means to fit a multivariate function to a given number of samples. Critically, the fitting should be predictive. After training a model on the gene expression profiling of some tumor and normal tissue samples, for instance, we hope that it can accurately determine if a new tissue sample is tumor or normal. More importantly, learning may also help us to discover the underlying biological way by fitting experimental biology data. For example, we may computationally determine the genetic markers related with cancer. In practice, learning techniques also have to be efficient so that we can deal with the flood of genetic data. Following the above principles, we have proposed several learning methods for gene finding and gene expression analysis, including predicting translation initiation sites in eukaryotic mRNAs with support vector machines and edit kernels, accurate and robust cancer classification with gene expression profiling, minimum entropy clustering method, and a general framework for biclustering gene expression data. Because the EM algorithm has been widely applied in computational biology and bioinformatics, we also develop a regularized EM algorithm that can effectively reduce the uncertainty of missing data. |
|
|
Book ReviewsPost a book review for this title
No reviews for this title. Be the first to post a review. |
More Bioinformatics BooksMore Bioinformatics Books ... |
| |
| |
|
|
|