Computational Inference STAT 440 / 840 CM 461 Instructor: Ali GhodsiAli Ghodsi Course Webpage: aghodsib/courses.

Computational InferenceComputational InferenceSTAT 440 / 840STAT 440 / 840

CM 461CM 461

Instructor: Ali Ghodsi

Course Webpage:http://www.math.uwaterloo.ca/~aghodsib/courses

ApplicationsApplications

• Computer Vision

• Speech Processing

• Machine Learning

• Molecular Biology

Computer VisionComputer Vision

N. Jojic and B.J. Frey, “ Learning flexible sprites in video layers”, CVPR 2001, (Video)

Artistic Painting Artistic Painting Style Translation Style Translation ((Unsupervised Unsupervised

Approach)Approach)CezanneCistern in the Park at Chateau Noir

Artistic Artistic painting painting

Texture TransferTexture Transfer

Model Representation Model Representation Probabilistic ModelProbabilistic Model

Image patches(output)

Image patches(input)

T. Transform.Filter

Romer Rosales

Speech ProcessingSpeech Processing((Denoising)Denoising)

Input signal (corrupted speech)

Denoising using a low pass filter

Denoising using Probabilistic Graphical Model

K. Achan, S. T. Roweis, A. Hertzmann, and B. J. Frey, 2004 A Segmental HMM for Speech Waveforms

Machine LearningMachine Learning(Spectral Clustering)(Spectral Clustering)

Machine LearningMachine Learning(Generative Models of Affinity Matrices(Generative Models of Affinity Matrices )

Molecular BiologyMolecular BiologyA REVISED VIEW OF THE MAMMALIAN LIBRARY OF

GENES (NATURE GENETICS, Aug 2005)

• Recent mammalian microarray experiments have detected widespread transcription and raised the possibility that there may be a large number of undiscovered multi-exon protein-coding genes. To explore this possibility, we hybridized unamplified, polyadenylation-selected samples from 37 mouse tissues to microarrays encompassing 1.14 million exon probes (see toy schematic on left). We analyzed these data using GenRate, a Bayesian algorithm that uses a genome-wide scoring function in a factor graph to infer genes. At a stringent exon false detection rate of 2.7%, GenRate detects 12,145 gene-length transcripts and confirms 81% of the 10,000 most highly-expressed known genes. Surprisingly, our analysis shows that most of the 155,839 exons detected by GenRate are associated with known genes, providing for the first time microarray-based evidence that the vast majority of multi-exon genes have already been discovered. GenRate also detects tens of thousands of potential new exons and reconciles discrepancies in current cDNA databases, by stitching novel transcribed regions into previously-annotated genes.