From castillo@myth.sdsu.edu Tue Oct 2 15:48:13 2007 From: castillo@myth.sdsu.edu (Jose Castillo) Date: Tue, 2 Oct 2007 07:48:13 -0700 Subject: [compsci] This Friday's Colloquium -> BAYESIAN MULTI-LDA FOR MATRIX FACTORIZATION Message-ID: Title: BAYESIAN MULTI-LDA FOR MATRIX FACTORIZATION Date: Friday, October 5th, 2007 Time: 3:30 PM Location: GMCS 214 Speaker: Ian Porteous Information & Computer Science University of California at Irvine Abstract: Matrix factorization algorithms like SVD and Non-negative matrix factorization (NMF) are arguably among the most widely used algorithms throughout machine learning and data mining. Applications of these techniques can be found in bioinformatics, robotics, computer vision, text analysis, information retrieval, collaborative filtering and so on. However, SVD and NMF do not assign probabilities to predictions and therefore can not provide uncertainty estimates. As an example a company wants to send gift cards to customers for items they are likely to want, but have not yet purchased. The company can run a matrix factorization algorithm on the joint product-customer matrix of ratings to find people who are likely to purchase item X, but the company will have no estimate of uncertainty for the predictions. Probabilistic models such as PLSA and their Bayesian extensions such as LDA which do assign probabilities to predictions have been proposed as text models in the bag-of-words representation. Furthermore, an extension of PLSA to the case of user recommendation systems has been proposed. However, these models treat customers and products differently. In particular they discover user communities but not product groups. We propose a symmetrized LDA model, which we call "Multi-LDA" which draws information from related products as well as related customers. Additionally, Multi-LDA is not limited to matrix factorization, but applies to tensor factorization as well. For example Multi-LDA could be applied to customer-product-date data. In addition to describing Multi-LDA, I will discuss how Multi-LDA relates to NMF and LDA and discuss some experimental results from customer-product ratings, customer-movie ratings, and handwritten digits. I will also outline a Nonparametric version, which is able to estimate the number of customer and product groups automatically. Host: Kristin Duncan For future events, please visit our web site at: http://www.csrc.sdsu.edu/csrc/events/colloquium/ ******************************************* Akmal Aulia PhD Student Computational Science Research Center San Diego State University Phone: 619-335-7187 Email: aaulia@sciences.sdsu.edu _______________________________________________ SDSU Computational Science Research Center Mailing List -------------------------------------------------------------------- myhosting.com - Premium Microsoft® Windows® and Linux web and application hosting - http://link.myhosting.com/myhosting _______________________________________________ SDSU Computational Science Research Center Mailing List -- Jose E. Castillo Ph.D. Director / Professor Computational Science Research Center 5500 Campanile Dr San Diego State University San Diego CA 92182-1245 619 5947205/3430, Fax 619-594-2459 castillo@myth.sdsu.edu http://www.csrc.sdsu.edu From aaulia@sciences.sdsu.edu Thu Oct 4 06:13:44 2007 From: aaulia@sciences.sdsu.edu (aaulia@sciences.sdsu.edu) Date: Thu, 4 Oct 2007 01:13:44 -0400 Subject: [compsci] *Reminder* This Friday's Colloquium -> BAYESIAN MULTI-LDA FOR MATRIX FACTORIZATION Message-ID: <380-22007104451344739@M2W030.mail2web.com> Title: BAYESIAN MULTI-LDA FOR MATRIX FACTORIZATION Date: Friday, October 5th, 2007 Time: 3:30 PM Location: GMCS 214 Speaker: Ian Porteous Information & Computer Science University of California at Irvine Abstract: Matrix factorization algorithms like SVD and Non-negative matrix factorization (NMF) are arguably among the most widely used algorithms throughout machine learning and data mining. Applications of these techniques can be found in bioinformatics, robotics, computer vision, text analysis, information retrieval, collaborative filtering and so on. However, SVD and NMF do not assign probabilities to predictions and therefore can not provide uncertainty estimates. As an example a company wants to send gift cards to customers for items they are likely to want, but have not yet purchased. The company can run a matrix factorization algorithm on the joint product-customer matrix of ratings to find people who are likely to purchase item X, but the company will have no estimate of uncertainty for the predictions. Probabilistic models such as PLSA and their Bayesian extensions such as LDA which do assign probabilities to predictions have been proposed as text models in the bag-of-words representation. Furthermore, an extension of PLSA to the case of user recommendation systems has been proposed. However, these models treat customers and products differently. In particular they discover user communities but not product groups. We propose a symmetrized LDA model, which we call "Multi-LDA" which draws information from related products as well as related customers. Additionally, Multi-LDA is not limited to matrix factorization, but applies to tensor factorization as well. For example Multi-LDA could be applied to customer-product-date data. In addition to describing Multi-LDA, I will discuss how Multi-LDA relates to NMF and LDA and discuss some experimental results from customer-product ratings, customer-movie ratings, and handwritten digits. I will also outline a Nonparametric version, which is able to estimate the number of customer and product groups automatically. Host: Kristin Duncan For future events, please visit our web site at: http://www.csrc.sdsu.edu/csrc/events/colloquium/ ******************************************* Akmal Aulia PhD Student Computational Science Research Center San Diego State University Phone: 619-335-7187 Email: aaulia@sciences.sdsu.edu _______________________________________________ SDSU Computational Science Research Center Mailing List -------------------------------------------------------------------- mail2web.com - Microsoft® Exchange solutions from a leading provider - http://link.mail2web.com/Business/Exchange From aaulia@sciences.sdsu.edu Sat Oct 6 23:16:57 2007 From: aaulia@sciences.sdsu.edu (aaulia@sciences.sdsu.edu) Date: Sat, 6 Oct 2007 18:16:57 -0400 Subject: [compsci] Next Friday's Colloquium -> BIOBRIDGE: LINKING THE SCIENTIFIC AND EDUCATIONAL COMMUNITIES Message-ID: <380-220071066221657196@M2W020.mail2web.com> Title: BIOBRIDGE: LINKING THE SCIENTIFIC AND EDUCATIONAL COMMUNITIES Date: Friday, October 12th, 2007 Time: 3:30 pm Location: GMCS 214 Speaker: Jeremy Babendure Biomedical Sciences University of California San Diego Abstract: The BioBridge California Partnership will address the national need for improved science inquiry curriculum to strengthen K-12 student engagement and achievement in the sciences. We propose to develop new curriculum with in-service teachers based on current research projects. Pedagogy used to develop the curriculum will be informed by current educational research, and modular units will be used by teachers to tailor curriculum for their students. Professional development workshops enable local teachers to master a variety of teaching strategies proposed for each unit. Host: Paul Paolini For future events, please visit our web site at: http://www.csrc.sdsu.edu/csrc/events/colloquium/ ******************************************* Akmal Aulia PhD Student Computational Science Research Center San Diego State University Phone: 619-335-7187 Email: aaulia@sciences.sdsu.edu _______________________________________________ SDSU Computational Science Research Center Mailing List -------------------------------------------------------------------- mail2web.com - Microsoft® Exchange solutions from a leading provider - http://link.mail2web.com/Business/Exchange From aaulia@sciences.sdsu.edu Tue Oct 9 14:50:51 2007 From: aaulia@sciences.sdsu.edu (aaulia@sciences.sdsu.edu) Date: Tue, 9 Oct 2007 09:50:51 -0400 Subject: [compsci] *Reminder* This Friday's Colloquium -> BIOBRIDGE: LINKING THE SCIENTIFIC AND EDUCATIONAL COMMUNITIES Message-ID: <380-220071029135051535@M2W028.mail2web.com> Title: BIOBRIDGE: LINKING THE SCIENTIFIC AND EDUCATIONAL COMMUNITIES Date: Friday, October 12th, 2007 Time: 3:30 pm Location: GMCS 214 Speaker: Jeremy Babendure Biomedical Sciences University of California San Diego Abstract: The BioBridge California Partnership will address the national need for improved science inquiry curriculum to strengthen K-12 student engagement and achievement in the sciences. We propose to develop new curriculum with in-service teachers based on current research projects. Pedagogy used to develop the curriculum will be informed by current educational research, and modular units will be used by teachers to tailor curriculum for their students. Professional development workshops enable local teachers to master a variety of teaching strategies proposed for each unit. Host: Paul Paolini For future events, please visit our web site at: http://www.csrc.sdsu.edu/csrc/events/colloquium/ ******************************************* Akmal Aulia PhD Student Computational Science Research Center San Diego State University Phone: 619-335-7187 Email: aaulia@sciences.sdsu.edu _______________________________________________ SDSU Computational Science Research Center Mailing List -------------------------------------------------------------------- mail2web.com – What can On Demand Business Solutions do for you? http://link.mail2web.com/Business/SharePoint From aaulia@sciences.sdsu.edu Tue Oct 16 00:17:20 2007 From: aaulia@sciences.sdsu.edu (aaulia@sciences.sdsu.edu) Date: Mon, 15 Oct 2007 19:17:20 -0400 Subject: [compsci] This Friday's Colloquium -> RESEARCH AT THE ROBOTICS AND MECHATRONICS LAB AT SDSU Message-ID: <380-2200710115231720356@M2W044.mail2web.com> Title: RESEARCH AT THE ROBOTICS AND MECHATRONICS LAB AT SDSU Date: Friday, October 19th, 2007 Time: 3:30 pm Location: GMCS 214 Speaker: Jingang Yi Department of Mechanical Engineering San Diego State University Abstract: In this talk, I will review some research activities at the Robotics and Mechatronics Lab at SDSU. In the first part of this talk, I will discuss some research activities in the vehicle and transportation area. Two subjects will be presented: safety-preserved tire/road interaction estimation and control, and macroscopic traffic flow stability. Tire/road friction characteristics determine the vehicle’s safety limits and need to be monitored and estimated on-line for vehicle safe operations. I will describe two different schemes for on-line estimation and control of the tire/road friction coefficient. An attractive property of these two approaches is that the estimates of the maximum friction coefficient and slip remain below their true values, which, in turn, guarantees safety of the vehicle spacing policies. I will also discuss some ongoing work on the development of a “smart tire” system. At traffic flow level, I will present some results on macroscopic traffic flow stability. A wavefront expansion technique is applied to establish a traffic flow stability criterion for both manual and adaptive cruise controlled vehicles. In the second part of this talk, I will present a design of vision-based motion planning and trajectory tracking control of an autonomous motorcycle. I will discuss how a vision vector space, a unitary vector set that represents local collision-free directions in the image coordinate system, is constructed for the highly dynamic requirements of motion planning of a motorcycle. I will also present a balance and trajectory tracking control algorithm for an autonomous motorcycle. The planning and control systems are validated by experimental and numerical testing based on a real prototype motorcycle system. HOST: Jose E. Castillo For future events, please visit our web site at: http://www.csrc.sdsu.edu/csrc/events/colloquium/ ******************************************* Akmal Aulia PhD Student Computational Science Research Center San Diego State University Phone: 619-335-7187 Email: aaulia@sciences.sdsu.edu _______________________________________________ SDSU Computational Science Research Center Mailing List -------------------------------------------------------------------- mail2web.com – What can On Demand Business Solutions do for you? http://link.mail2web.com/Business/SharePoint From aaulia@sciences.sdsu.edu Thu Oct 18 07:38:57 2007 From: aaulia@sciences.sdsu.edu (aaulia@sciences.sdsu.edu) Date: Thu, 18 Oct 2007 02:38:57 -0400 Subject: [compsci] *Reminder* This Friday's Colloquium -> RESEARCH AT THE ROBOTICS AND MECHATRONICS LAB AT SDSU Message-ID: <380-220071041863857492@M2W032.mail2web.com> Title: RESEARCH AT THE ROBOTICS AND MECHATRONICS LAB AT SDSU Date: Friday, October 19th, 2007 Time: 3:30 pm Location: GMCS 214 Speaker: Jingang Yi Department of Mechanical Engineering San Diego State University Abstract: In this talk, I will review some research activities at the Robotics and Mechatronics Lab at SDSU. In the first part of this talk, I will discuss some research activities in the vehicle and transportation area. Two subjects will be presented: safety-preserved tire/road interaction estimation and control, and macroscopic traffic flow stability. Tire/road friction characteristics determine the vehicle’s safety limits and need to be monitored and estimated on-line for vehicle safe operations. I will describe two different schemes for on-line estimation and control of the tire/road friction coefficient. An attractive property of these two approaches is that the estimates of the maximum friction coefficient and slip remain below their true values, which, in turn, guarantees safety of the vehicle spacing policies. I will also discuss some ongoing work on the development of a “smart tire” system. At traffic flow level, I will present some results on macroscopic traffic flow stability. A wavefront expansion technique is applied to establish a traffic flow stability criterion for both manual and adaptive cruise controlled vehicles. In the second part of this talk, I will present a design of vision-based motion planning and trajectory tracking control of an autonomous motorcycle. I will discuss how a vision vector space, a unitary vector set that represents local collision-free directions in the image coordinate system, is constructed for the highly dynamic requirements of motion planning of a motorcycle. I will also present a balance and trajectory tracking control algorithm for an autonomous motorcycle. The planning and control systems are validated by experimental and numerical testing based on a real prototype motorcycle system. HOST: Jose E. Castillo For future events, please visit our web site at: http://www.csrc.sdsu.edu/csrc/events/colloquium/ ******************************************* Akmal Aulia PhD Student Computational Science Research Center San Diego State University Phone: 619-335-7187 Email: aaulia@sciences.sdsu.edu _______________________________________________ SDSU Computational Science Research Center Mailing List -------------------------------------------------------------------- mail2web.com – What can On Demand Business Solutions do for you? http://link.mail2web.com/Business/SharePoint From aaulia@sciences.sdsu.edu Sun Oct 21 22:18:55 2007 From: aaulia@sciences.sdsu.edu (aaulia@sciences.sdsu.edu) Date: Sun, 21 Oct 2007 17:18:55 -0400 Subject: [compsci] Colloquium: DESIGN OF CHEMICAL AND BIOLOGICAL DATABASES & VIRTUAL SCREENING Message-ID: <380-2200710021211855419@M2W037.mail2web.com> Title: DESIGN OF CHEMICAL AND BIOLOGICAL DATABASES & VIRTUAL SCREENING Date: Friday, October 26th, 2007 Time: 3:30 pm Location: GMCS 214 Speaker: Srinivas Reddy Alla Chemistry and Biochemistry Department, California State University, San Marcos; CA, USA Molecular Modeling Group, Indian Institute of Chemical Technology, Hyderabad, INDIA Abstract: The approach to understand the chemical and biological problems has taken a new turn in the post-genomic era, due to the availability of vast amount of related biological and chemical data. To unravel the useful information from this data, use of computational techniques in bioinformatics and chemoinformatics have become an integral part of modern drug discovery, design, and development. This is especially true for the identification and validation of drug targets as well as for the screening and design of potential bioactive substances. The present talk will introduce the available database design methodologies, and the discussion on the developed in-house databases Viz. SiRNA (siRNA database for Oncogenes), CAD (Cation-Aromatic Interaction Database), PLID (Protein Ligand Interaction Database) Later part of the talk will introduce the virtual screening approaches available in the drug discovery process. HOST: Jose E. Castillo For future events, please visit our web site at: http://www.csrc.sdsu.edu/csrc/events/colloquium/ ******************************************* Akmal Aulia PhD Student Computational Science Research Center San Diego State University Phone: 619-335-7187 Email: aaulia@sciences.sdsu.edu _______________________________________________ SDSU Computational Science Research Center Mailing List -------------------------------------------------------------------- myhosting.com - Premium Microsoft® Windows® and Linux web and application hosting - http://link.myhosting.com/myhosting From aaulia@sciences.sdsu.edu Tue Oct 23 00:45:40 2007 From: aaulia@sciences.sdsu.edu (aaulia@sciences.sdsu.edu) Date: Mon, 22 Oct 2007 19:45:40 -0400 Subject: [compsci] SDSU Aerospace Engineering Seminar Message-ID: <380-2200710122234540645@M2W022.mail2web.com> I would like to invite you to attend the following seminar http://www.engineering.sdsu.edu/aerospace/seminars.html Thursday at 1400 Regards, Gustaaf (Guus) Jacobs, Assistant Professor Department of Aerospace Engineering San Diego State University Tel: 619-594-4046 Fax: 619-594-6005 http://attila.sdsu.edu/~jacobs/ -------------------------------------------------------------------- myhosting.com - Premium Microsoft® Windows® and Linux web and application hosting - http://link.myhosting.com/myhosting From aaulia@sciences.sdsu.edu Fri Oct 26 05:18:20 2007 From: aaulia@sciences.sdsu.edu (aaulia@sciences.sdsu.edu) Date: Fri, 26 Oct 2007 00:18:20 -0400 Subject: [compsci] CANCELLATION: CSRC colloquium and SDSU Aerospace Engineering seminar Message-ID: <380-220071052641820265@M2W042.mail2web.com> The recently posted CSRC Friday's Colloquium and Dr. Jacobs Gustaaf's SDSU Aerospace Engineering seminar this Thursday has been cancelled, and will be rescheduled. Best regards, Akmal Aulia Attached: Hi: The seminar on thursday has been canceled and will be rescheduled. Best wishes, Guus ----------------------------------------------- Gustaaf (Guus) Jacobs, Assistant Professor Department of Aerospace Engineering San Diego State University Tel: 619-594-4046 Fax: 619-594-6005 http://attila.sdsu.edu/~jacobs/ ----------------------------------------------- -------------------------------------------------------------------- myhosting.com - Premium Microsoft® Windows® and Linux web and application hosting - http://link.myhosting.com/myhosting From aaulia@sciences.sdsu.edu Sun Oct 28 03:52:52 2007 From: aaulia@sciences.sdsu.edu (aaulia@sciences.sdsu.edu) Date: Sat, 27 Oct 2007 22:52:52 -0400 Subject: [compsci] Colloquium: DESIGN OF CHEMICAL AND BIOLOGICAL DATABASES & VIRTUAL SCREENING Message-ID: <380-220071002825252411@M2W029.mail2web.com> Title: DESIGN OF CHEMICAL AND BIOLOGICAL DATABASES & VIRTUAL SCREENING Date: Friday, November 2nd, 2007 Time: 3:30 pm Location: GMCS 214 Speaker: Srinivas Reddy Alla Chemistry and Biochemistry Department, California State University, San Marcos; CA, USA Molecular Modeling Group, Indian Institute of Chemical Technology, Hyderabad, INDIA Abstract: The approach to understand the chemical and biological problems has taken a new turn in the post-genomic era, due to the availability of vast amount of related biological and chemical data. To unravel the useful information from this data, use of computational techniques in bioinformatics and chemoinformatics have become an integral part of modern drug discovery, design, and development. This is especially true for the identification and validation of drug targets as well as for the screening and design of potential bioactive substances. The present talk will introduce the available database design methodologies, and the discussion on the developed in-house databases Viz. SiRNA (siRNA database for Oncogenes), CAD (Cation-Aromatic Interaction Database), PLID (Protein Ligand Interaction Database) Later part of the talk will introduce the virtual screening approaches available in the drug discovery process. HOST: Sunil Kumar For future events, please visit our web site at: http://www.csrc.sdsu.edu/csrc/events/colloquium/ ******************************************* Akmal Aulia PhD Student Computational Science Research Center San Diego State University Phone: 619-335-7187 Email: aaulia@sciences.sdsu.edu _______________________________________________ SDSU Computational Science Research Center Mailing List -------------------------------------------------------------------- mail2web.com - Microsoft® Exchange solutions from a leading provider - http://link.mail2web.com/Business/Exchange From aaulia@sciences.sdsu.edu Wed Oct 31 19:54:25 2007 From: aaulia@sciences.sdsu.edu (aaulia@sciences.sdsu.edu) Date: Wed, 31 Oct 2007 14:54:25 -0400 Subject: [compsci] *Reminder* Colloquium: DESIGN OF CHEMICAL AND BIOLOGICAL DATABASES & VIRTUAL SCREENING Message-ID: <380-2200710331185425345@M2W013.mail2web.com> Title: DESIGN OF CHEMICAL AND BIOLOGICAL DATABASES & VIRTUAL SCREENING Date: Friday, November 2nd, 2007 Time: 3:30 pm Location: GMCS 214 Speaker: Srinivas Reddy Alla Chemistry and Biochemistry Department, Cal. State Univ., San Marcos; CA, USA Molecular Modeling Group, Indian Institute of Chem. Tech., Hyderabad, INDIA Abstract: The approach to understand the chemical and biological problems has taken a new turn in the post-genomic era, due to the availability of vast amount of related biological and chemical data. To unravel the useful information from this data, use of computational techniques in bioinformatics and chemoinformatics have become an integral part of modern drug discovery, design, and development. This is especially true for the identification and validation of drug targets as well as for the screening and design of potential bioactive substances. The present talk will introduce the available database design methodologies, and the discussion on the developed in-house databases Viz. SiRNA (siRNA database for Oncogenes), CAD (Cation-Aromatic Interaction Database), PLID (Protein Ligand Interaction Database) Later part of the talk will introduce the virtual screening approaches available in the drug discovery process. HOST: Sunil Kumar For future events, please visit our web site at: http://www.csrc.sdsu.edu/csrc/events/colloquium/ ******************************************* Akmal Aulia PhD Student Computational Science Research Center San Diego State University Phone: 619-335-7187 Email: aaulia@sciences.sdsu.edu _______________________________________________ SDSU Computational Science Research Center Mailing List -------------------------------------------------------------------- mail2web.com – What can On Demand Business Solutions do for you? http://link.mail2web.com/Business/SharePoint