FIRE271: Capital One Machine Learning - Fall 2018
FIRE271 is the third and final semester First-Year Innovation & Research Experience (FIRE) course for students enrolling in the FIRE Capital One Machine Learning Research Stream. The goal of the research and innovation stream is to inspire students to produce new and original work in the discipline of machine learning. In this course, students will continue their exploration of the exciting field of machine learning.
FIRE271 is a collaborative research engagement focusing on the development and deployment of Machine Learning. Machine learning is a subdiscipline of computer science focused on the capacities for computer systems to learn without being purposefully programmed to perform specific tasks. These capacities derive from pattern recognition and artificial intelligence (AI). The Capital One Machine Learning research stream will focus on using machine learning to develop algorithms involved in predictive data analytics using both supervised and unsupervised approaches.
This semester will focus on the process of independent research, including collaboration with peers, communication of ideas, troubleshooting unexpected outcomes, and discipline-specific methodologies. Scheduled class meetings will focus on the discussion of primary literature, troubleshooting research issues, and continual review of individual and group research progress. The course requires students to commit 8 additional hours of independent research per week (6 hours in the research space and 2 hours off-site).
The goal of the research and innovation stream is to inspire students to produce new and original work in the discipline of machine learning. In 2018, the stream will explore the state-of-art techniques in machine learning, data analytics and cybersecurity using recently developed perspectives, technologies and applications.
Machine Learning for Computer Vision
It is estimated that by 2019, 84% of the world's Internet traffic will be visual. Visual object recognition is enabling innovative systems like semantic-based image retrieval, autonomous unmanned systems, biomedical image applications. Deep learning added a huge boost to the already rapidly developing field of computer vision. However, the deep learning models these systems rely on can be difficult to design, train, and evaluate. With these pressing challenges, can we design an algorithm that efficiently learns to model deep neural networks for computer vision problems?
- Detecting Driver Drowsiness and Attentiveness Through Facial Recognition
- Electronic Access Control System Using Facial Recognition, Gait Analysis, and Global Scene Understanding
- What’s This? Landmark Detection in DC
- Predicting Bee Behavior
Machine Learning for Natural Language Processing
Every day, we are awash with text, from blogs, tweets, news, books, papers, and increasingly text from spoken utterances. Working with natural language is challenging given the inherent ambiguity and flexibility of human language. The recent development of encoder-decoder recurrent neural network architectures achieved state-of-the-art results compared to classical rule-based systems. Although effective, the neural machine translation systems still suffer some issues, such as slower training and inference speed, ineffectiveness in dealing with rare words, and sometimes failure to translate all words in the source sentence. With these issues, can we design a model that more completely represent linguistic ideas of a given text?
- Multimodal Sentiment Analysis for Visual, Textual, and Aural Content
- Improving 9-1-1 Operating Efficiency Using Natural Language Processing
Machine Learning for Cybersecurity and Fraud Detection
Detection of fraudulent activity in commercial transactions presents a significant opportunity potentially worth billions of dollars per year. Creating automated systems that can detect fraud is a natural machine learning problem. However, shifts in user behavior over time can create biases in the predictions that they make. A cybercriminal could potentially generate adversarial sequences to avoid flagging future transactions as fraudulent. Given historical transaction data, can we design an algorithm that learns to robustly detect fraudulent transactions?
- Improving Current Techniques in Deep Learning for Time Series Analysis in the Context of Sales Forecasting
- Deep Learning Approach to Credit Card Fraud Detection Using Time Series Data
- AI Versus the Complexity of Malicious Code
Here's a photo of who we are:
Capital One offers a broad array of financial products and services to consumers, small businesses and commercial clients in the U.S., Canada and the UK. Capital One is a major partner in the design and creation of the FIRE stream Capital One Machine Learning .
Google Cloud Platform (GCP) is a suite of public cloud computing services offered by Google. The platform includes a range of hosted services for compute, storage and application development that run on Google hardware. GCP is a sponsor in providing cloud platform credits for the FIRE stream Capital One Machine Learning.
Amazon Web Services (AWS) is a secure cloud services platform, offering compute power, database storage, content delivery and other functionality to help businesses scale and grow. AWS is a sponsor in providing cloud platform credits for the FIRE stream Capital One Machine Learning.
IBM Cloud is a cloud platform that offers a choice of scalable and flexible resources in one consistent experience. Bringing together APIs and services, IBM Cloud offers a rich and continuously expanding ecosystem of services to accelerate the pace of innovation. IBM Cloud is a sponsor in providing cloud platform credits for the FIRE stream Capital One Machine Learning.