Topic: An Examination of Adversarial Machine Learning on Securing Connected and Autonomous Vehicles LOCATION: Lester W. Cory Conference Room, Science & Engineering Building (SENG), Room 213A Abstract: Connected and Autonomous Vehicle (CAV), a.k.a. driverless car, provides a solution to safety and efficiency in surface transportation. Yet, CAV technology introduces new security risks beyond the standard vehicle vulnerabilities. Machine Learning (ML) promises to secure CAV by utilizing telemetry data to validate legitimacy among CAVs as a method of Misbehavior Detection. However, most ML models base a misassumption of a benign environment, where the data during testing and training keep an equivalent distribution, which makes ML models vulnerable to adversarial perturbations. Furthermore, the lack of knowledge in CAV behavior opens new attack surfaces for adversaries to invalidate ML models. This thesis examines an attackers' mindset to fool a ML system for Misbehavior Detection in CAV; thereby CAV defense strategies become robust to adversarial ML. Three software packages, OMNET++, SUMO, and VEINS, integrate to simulate CAV. The dataset for ML, comes from the Vehicle Reference Misbehavior Dataset (VeReMi), which this work expands with new attacks. Progressively, the work evaluates the accuracy of K-Nearest Neighbors (KNN) and Random Forest (RF) for ML algorithms and that of Logistic Regression Neural Network (LRNN) and Recurrent Neural Network (RNN) with Long-Short-Term-Memory (LSTM) for Deep Learning (DL) algorithms, all coded in Python. A Random Walk movement theory generates synthetic data following a normal distribution to further evaluate the robustness of the ML/DL models, i.e. the effectiveness of spoofed data from the attackers' point of view. The work reaches a conclusion that DL models perform better than ML models in most cases. This work contributes to ML Misbehavior Detection in CAV, exposing the challenges posed by adversarial ML and recommending the use of DL over ML against adversaries. As a result, it strengthens CAV defense with robust DL models. Future work will solve the limitations of DL such as the need for large quantity of data and balanced classes of data. To correct unbalanced classes, synthetic data generation can use specific python packages such as SMOTE. Undersampling or oversampling the minority or majority classes, respectively, could be another solution. Pairing Feature Analysis, such as Exploratory Data Analysis, to identify the most relevant features would train the models effectively and accurately. Other future work includes engineering the features from the datasets as well as using the Real-World Datasets from CAV pilot cities. NOTE: All ECE Graduate Students are ENCOURAGED to attend. All interested parties are invited to attend. Open to the public. Advisor: Dr. Hong Liu Committee Members: Dr. Liudong Xing and Dr. Benjamin Viall, Department of Electrical & Computer Engineering, UMASS Dartmouth *For further information, please contact Dr. Hong Liu at 508.999.8514, or via email at firstname.lastname@example.org.
The Boivin Center for French Language and Culture presents The Seine Comes to America, featuring New York Times contributing writer and best-selling author, Elaine Sciolino. The event will take place on campus in the Charlton Business College Auditorium (CCB149) at 7 PM. A Q&A, book signing and light reception will follow the lecture which is free and open to the public. Parking is available in lot 13. For further information regarding this program, please contact Maria Sanguinetti at email@example.com or by calling 508-991-5096.
Italian Film Series: Suspiria (2018) Date: Tuesday, December 10, 2019 Time: 7:00 PM Place: LARTS 117 Description: Join us for our 5th Annual Italian Film Series at UMass Dartmouth, co-sponsored by the Foreign Literature and Languages Department and the History Department. This semester, we will be delving into various Italian films in the horror genre. For our fourth and final film this semester, we will show Suspiria (2018): It is a remake of the 1977 Dario Argento film of the same title. A darkness swirls at the center of a world-renowned dance company, one that will engulf the artistic director, an ambitious young dancer, and a grieving psychotherapist. Some will succumb to the nightmare. Others will finally wake up. (IMDB). Free admission. All films have English subtitles. Refreshments will be served. Any questions may be directed to Prof. Rose Facchini at firstname.lastname@example.org or Prof. Matthew Sneider at email@example.com.
The Psychology Department will be holding its annual Fall Semester Research Conference. Starting at 4PM there will be two research talks: Infants can use linguistic information to infer others psychological states by Dr. Hyun-joo Song of the Department of Psychology, Yonsei University (Korea) and Graduate School of Education, Harvard University, and Developing novel strategies for overcoming barriers to physical activity among adults with obesity by Dr. Leah Schumacher of the Weight Control and Diabetes Research Center, The Miriam Hospital/Alpert Medical School of Brown University. At 5:30PM Psychology students will present posters on research projects within the areas of Developmental, Social/Personality, Physiological, and Clinical Psychology. For more information please contact Dr. Robin Locke-Arkerson at firstname.lastname@example.org or 508-910-6955. All are welcome to attend. Light refreshments will be served.
Join us on December 12th for an extraordinary evening with Dolores Hirschmann. Dolores is an experienced entrepreneur, coach, speaker and organizer of TedX talks. Dolores will be sharing a talk entitled: 18 MINUTES OF IMPACT: Moving from delivering information to engaging your audience. This topic is designed for decision-makers, C-level executives, sales teams, and entrepreneurs. Participants will learn a step-by-step framework that will help them create presentations that engage and move their audiences to action. They will understand that every time they are in front of an audience they are responsible for an EMOTIONAL experience, and they will learn how to craft the message AND the experience into a seamless talk that uses both relevant storytelling, data, and case studies. KEY TAKEAWAYS: 1) Participants will craft a talk they can present to the world 2) Participants will learn how to put presentations together easily and the best practices in effective presentations including the use of stories and how to integrate data and statistics while keeping the engagement 3) Participants will discover that the first 60 seconds of a presentation are the most important to ensure the audience is engaged and invested in what they have to say About Dolores Dolores is an internationally recognized STRATEGIST & COACH. She helps clients clarify their "idea worth sharing," design their communication strategies, and implement business growth systems. Her clients become speakers and authors and take their message to larger audiences like TEDx and beyond. She works through group coaching, workshops, one on one coaching, as well as public speaking. Dolores is a writer, TEDx Organizer, and participant in TED conferences. She is a CTI certified and ICF accredited coach and has a business degree from the Universidad de San Andres, Argentina. Originally from Buenos Aires, Dolores speaks fluent Spanish, English, and French and lives in Dartmouth, MA with her husband and four children.
TITLE : Electro-mechanical Response of Multi-functional Jute/Epoxy Composites DATE: Friday, December 13, 2019 TIME: 10 AM - 12PM LOCATION: Textiles - 101E Abstract: A comprehensive investigation is performed in this study to determine the electro-mechanical response for damage detection in novel multi-functional natural fiber composites under various mechanical loads. Both experiments and theoretical studies are performed in this research. A jute/epoxy composite system was fabricated by embedding multi-wall carbon nanotubes (MWCNTs) in the matrix and reinforcing short carbon fibers (CFs) between the jute laminates. The MWCNTs are embedded in the epoxy matrix using a combination of ultrasonication and shear mixing. The CFs are positioned along the through-thickness direction on the jute fiber laminates using UMass Dartmouth's patented 'wet flocking' technique. The embedded MWCNTs in the matrix and reinforced CFs between the laminates generate a three-dimensional electrically conductive network in the composite. Four process variables such as weight percentage of MWCNTs (0.1 and 0.025), flock length (150Î¼m and 350Î¼m), flock density (500, 1000, 1500, and 2000 fibers/mm2), and two jute fiber orientations ((0-0-0-0)T and (0-90-0-90)T) are considered in this investigation for all experiments. The change in the electrical network of the composite associated with the mechanical loading is determined by the four-circumferential ring probes (FCRP) technique. In this study, first, a parametric study is performed to investigate the electrical resistivity of various composites with the above-listed process variables. To have a complete understanding of damage detection in multi-functional composites, the electro-mechanical response of the composites is then investigated under interlaminar shear and fracture loading conditions. Since composites have curved structures in real-life situations, an experimental study is also performed to determine the electro-bending response of curved composites under flexural loading. Moreover, a theoretical model was developed to investigate the electro-mechanical responses of CNTs embedded epoxy under shear loading conditions. A parametric study was performed, and the electrical response was compared with that of experimental findings. Advisor: Dr. Vijaya Chalivendra, Department of Mechanical Engineering Committee Members: Dr. Jun Li, Department of Mechanical Engineering Dr. Yong Kim, Department of Bioengineering Dr. Jay Wang, Department of Physics All EAS students are encouraged to attend and all interested parties invited. For further information, please contact Dr. Vijaya Chalivendra by email at email@example.com