In 2010 alone, National Institutes of Health (NIH) spent $10.7 Billion (one-third of funds) on clinical trials research while pharmaceutical companies spent $32.5 Billion (70% of overall). However, we know little about the choices in co-funding and the impact generated by the co-funded projects. Every funding agency is targeted towards solving specific problems, but how inter-locked are these problems and how interconnected are the agencies? As an initial effort towards this, we explore the structure of a co-funding network, where a funding agency represents a node and an edge signifies co-funding of a project. Our core data on clinical trials includes all papers that cite and receive citations by the Cochrane Database of Systemic Reviews, a prominent journal that reviews clinical studies. From these papers, we find communities of co-funders that are guided by similar funding objectives and the primary region of operation. Next, to quantify success in funding, we use a g-index type metric that considers the average citations received by the top g most cited papers, which also indicates efficiency in funding by an agency. We find that as a funder, seeking multiple, direct connections with various dis-connected funders is more productive rather than being part of a densely inter-connected network of co-funders. This type of local network structure likely reduces redundancy in information sharing and enables efficient exchange of resources, ideas, and money. While this is not a causal relationship, we theorize that, co-funding strategies, when carefully mediated can allow a funder to achieve higher social capital which then leads to higher research success for clinical investment.

The unlimited potential uses of data, coupled with the power of artificial intelligence to find and learn patterns has the ability to transform our world and solve humanities problems. As a result of the abundance of financial data, there is a new opportunity to create models which can learn from patterns and predict future price movements. The goal of this project is to create a portfolio with pure alpha. A portfolio which will produce a consistent stream of returns with little risk. Using data analysis, the portfolio will be constructed with thirty to fifty uncorrelated assets, with the hypothesis that the uncorrelated assets will remove risk due to both general market volatility as well as movements from other assets in the portfolio. Machine learning and artificial intelligence techniques will be incorporated to design an algorithm which will continually learn from itself and optimize the weightings of the assets in the portfolio. To ensure optimal returns, I will use back testing to see how the algorithm would have performed historically in different market conditions, and how the machine would re-learn from past observations to improve in future out of sample testing. While predicting the future is difficult by humans, using machine learning techniques where self-learning is continuous, allows for faster predictions with less lag time and bias than humans are capable of.

Task prioritization, or choosing which order to do tasks, is an essential skill to possess. However, there are no definitive answers to the question of “how” we prioritize tasks. There are numerous factors that are considered in task prioritization, including a vast array of queuing rules (FIFO or First In First Out, EPT or Earliest Processing Time, etc.) However, many of these systematic methods of prioritization don’t account for task complexity. Another common way to prioritize tasks, which will be the focus of this study, is finding the task that seems the closest in time. This perception of Temporal Distance does address task complexity because we balance task complexity and due date when deciding which task is closer or farther away in time. Though individual Temporal Distance Perception is a widely researched topic, there has not been much research conducted on application in a team setting. Prioritizing one’s tasks individually is difficult, but it gets much harder when others’ schedules need to be factored in as well as an individual’s schedule. In addition to this issue of schedule navigation, this research will take place in a multidisciplinary setting, so teammates will not be able to know in detail how long a teammate’s task will take. This too makes it more difficult to choose which task comes first. My research question is this: How does Temporal Distance Perception in a Multidisciplinary Setting affect Task Prioritization in Teams? I am conducting a synthesis of literatures about Task Prioritization, Multidisciplinary Teamwork, and Temporal Distance Perception, as well as creating a data model based off the synthesis. I was sourced the results of team building exercises from NASA’s Human Research Program, which include a variety of real-world examples of how astronaut teams prioritize tasks.

Internet of Things (IoT) devices are increasingly integrated into new construction and legacy buildings. However, the insertion of IoT into buildings increases their cybersecurity vulnerability. While IoT security issues are ubiquitous across devices, a foundational issue for cybersecurity in buildings is the organizational friction due to the increasing integration between Information Technology (IT) and Operational Technology (OT), called IT/OT convergence. These two technological systems are embedded in the siloed institutional histories and work practices that historically managed these technologies: Operations and Management (O&M) and IT. This siloed evolution generated different organizational structures and cultures that facilitate a responsibility vacuum in organizations around IoT cybersecurity. Industry stakeholders suggest that stronger collaboration between O&M and IT could mitigate these challenges. However, improving collaboration requires multi-disciplinary knowledge work and task coordination that fits the policy and organizational contexts of O&M and IT professionals. This project seeks to identify different interactions between macro-level legal policies and formal organizational policies and procedures, and how micro-level daily work practices support the knowledge work and task coordination needed to improve O&M and IT collaboration around IoT cybersecurity. To tackle this subject, our team is undertaking a multi-method qualitative study of three parts: (1) ethnographic study of IOT cybersecurity practices by IT and O&M departments at three small, medium, and large universities in Washington state: UW Bothell, Western Washington University, and UW; (2) 25 expert interviews with IT and O&M cybersecurity professionals working in universities across the United States; (3) 15 case studies of universities in the Pacific Northwest region. The data collection methods for these three parts are observations in the workplace, interviews, and document analysis.The research being presented here will consist of two master inventories of (1) policies addressing cybersecurity, IoT device security, data privacy; and (2) smart city initiatives in the United States.