Fall Dining Hours
.jpg)
As the incoming chair of Visual Arts, I would like to invite you to a relaxed and friendly reception for Adjuncts on Thursday, September 6 from 12noon to 1:30pm in Fine Arts 342. I want to take the opportunity to meet and talk with each and every one of you and to introduce you to your peers and other Visual Arts faculty and staff. We appreciate all that you do for us and hope that your semester will be a satisfying and productive one. I look forward to seeing you there.
Please RSVP to sharda1@umbc.edu ]]>
In the last decade, statistical learning, which is the concept of using algorithms to identify patterns and/or make predictions based on input data sets, has received increased interest due to its potential to answer diverse questions in various industries such as finance, business, and health. In this talk, I will introduce the two most fundamental methods of statistical learning that are applicable to both data and computer science: regression and classification. Real world examples will be provided to highlight the differences and similarities between these two methods and place them into appropriate contexts.
Dr. Ergun Simsek earned his PhD from Duke University in 2006. He is the co-director of the Computer Science and Information Technologies Department at the Bahcesehir University’s newly established DC campus (BAU DC). He joined Bahcesehir University’s Electrical and Electronics Engineering Department and later was promoted to associate professor. Upon moving back to the United States, he spent six years at the George Washington University’s Department of Electrical & Computer Engineering before returning to BAU at their DC campus. Dr. Simsek’s private sector experience includes working for Schlumberger Doll Research (SDR) Center’s Math and Modeling Department as a post-doctoral research associate, where he helped develop new numerical techniques for various oil and gas industry applications. He continues researching how to solve emerging engineering problems through efficient and robust computational techniques.
Host: Dr. Richard Forno (*protected email*)
The post talk: Statistical Learning 101: Regression vs. Classification, 11:30 9/5 appeared first on Department of Computer Science and Electrical Engineering.
]]>**from Connor Strang, UMBC Robosub team**
The UMBC Robosub team recently competed for the first time in the International Robosub Competition in San Diego, California from July 30th to August 5th. The Robosub competition challenges teams to develop Underwater Autonomous Vehicles (AUVs) capable of maneuvering underwater, dropping markers, firing torpedoes, locating acoustic pingers and interacting with field elements designed around the casino theme for the year. The three members of the team who attended joined approximately 500 students and mentors across 47 teams and 10 countries in a cooperative effort to push the limits of AUV technology.
UMBC capstone students have been working diligently for the past four years to develop the structure for the robot. This year,
a cross-disciplinary group initially composed of Retriever Robotics members picked up the project and added the sensors and other hardware necessary to make the vehicle ready to compete.
In particular, the team integrated depth and orientation sensors, experimented with vision systems, developed software for motion control, and plugged the leaks. At the time the robot was shipped out o the competition, there was only one major part missing – a kill witch. The circuit for this was fabricated and installed from a hotel room over the course of the competition week. The lack of a kill witch led to the loss of valuable testing time early in the week as vehicles without this critical component were not allowed to operate in the main pool. As a result, the team had to wait three days to discover that their vision system was severely crippled by the unexpectedly low water quality.
Despite these setbacks, the team managed to qualify, perform adequately in the semi final mission using only feedback from the orientation sensor to hold course, and even get a chance at a
finals slot. Unfortunately, a minor issue in the mission code was not repaired in time to compete in the time slot. Despite these frustrations, the team’s ability to overcome challenges and their eagerness to
aid other teams led them to receive the “Best First Time Entry” award.
Using the experience gained in the competition and connecting with teams and staff, UMBC Robosub is in the process of developing several improvements to be incorporated into a new, more capable, design. In particular, we envision a strong partnership with IEEE members to develop active and passive sonar systems, and hope to recruit new members to help design and manufacture a power-
efficient and precise buoyancy control system.
Finally, UMBC Robosub would like to thank the Dean’s Office, with the Computer, Electrical, and Mechanical Engineering departments for their generous funding this year, and IEEE for partnering with us in that endeavor. In particular, we would like to thank BAE Systems, Inc. for their continued support and generosity over the years. We greatly appreciate the backing that allowed the members of UMBC Robosub the opportunity to develop a relationship with the amazing teams at the competition and share the wonder of autonomous robotics with the world.
Connor Strang
UMBC Robosub