Artificial intelligence architectures for pharmaceutical drug discovery
Pratik Shah, Francis Kendall (Roche) | E15-318
We are entering an era of exponential growth in applications of new artificial intelligence (AI) techniques to biological and clinical research problems. Research and clinical institutions generate large amounts of in vitro and in vivo data in the form of numbers and images. Complex datasets are also created from next-generation sequencing, transcriptomics, proteomics, metabolomics and single-cell experiments. High-throughput screens geared toward small-molecule discovery require manual processing and in-silico prediction of hits and potential targets and mechanism, while in vivo experiments and clinical trials are expensive and lack learning and predictions gleaned from past experimental successes and failures. All these processes run in parallel and generate large amounts of data that live in their respective silos, often making them untenable to sharing and intuitive analyses across the organization. In addition, data are captured and stored in formats that are untenable for machine learning (ML) or deep neural networks (DNN). In this workshop, we will discuss:
1) new algorithms for automated structuring of raw biological and clinical data for input into ML and DNN classifiers by bioinformatics and data science professionals;
2) new models and emerging DNN architectures for classification of multimodal clinical and biological datasets;
3) how do we build a horizontal pharma/bio data platform that all Media Lab member companies can contribute to and get value from?
Jifei Ou, Jie Qi, Artem Dementyev | E14-240
This workshop will explore how to foster innovation by bringing designers, hackers, and engineers to the manufacturing process. Instead of focusing on innovation with end products in mind, we would like to emphasize and explore the aspect of "processes," where the existing mass production pipeline is hacked, modified, or optimized.
The workshop will start with a presentation about our experience of working with Shenzhen factories to scale up research projects here at the Lab. We will also have a discussion about what we're calling "hacking the factories," a new path for collaboration between researchers and members.
Immersion teams: a tool to understand team dynamics from email interactions
César Hidalgo, Catherine D'Ignazio, Jingxian Zhang, Xiaojiao Chen | E14-393
The Collective Learning group (formerly Macro Connections) at the MIT Media Lab is building Immersion Teams: a tool that visualizes how teams and organizations are interconnected through their email communications. In this session we will explore the potential of this tool together with Media Lab member companies, and get feedback on the features people would like to see implemented in the tool.
Meet the startups
Habib Haddad | E14-3rd Floor Atrium
12 MIT Media Lab spin-offs will present their work and then join a moderated discussion with participants around the best ways for member companies and startups to work together. Participating startups: BioBright, Figur8, Lumii, OpenAg, Pienso, Soofa, Sourcemap, Tulip, Twine Health, Waylens, and Wise Systems.
Personalized machine learning for health and wellbeing
Rosalind Picard, Oggi Rudovic | E15-359
There is a need, now more than ever, for personalized models and applications that can efficiently leverage big data and adapt to target individuals, especially in smart healthcare and medicine. The main challenge in personalized modeling is how to use machine learning to build personalized models that work well for a target (risk) group and/or individual—thus, optimizing outcomes for each person, and not the average outcome for a group. In this workshop, we will introduce the main challenges in Personalized Machine Learning from human wellbeing and health data perspectives, and we'll discuss future directions and applications of this emerging learning approach.
Extended Reality SIG: learning, visualization, creativity, and collaboration in VR and AR
Mike Bove, Pattie Maes, Edward Boyden, Scott Greenwald, Dan Novy | E15-341
The workshop will introduce a new special interest group which aims to bring together multiple Media Lab research groups and member companies to advance research on virtual and augmented reality. We are specifically interested in the opportunities for extended reality technologies to radically impact learning as well as creative and collaborative work. We expect several industries to benefit, including entertainment, scientific research, design, training, and education. The SIG will have two foci: First, to both build novel prototypes to evaluate the potential of these new technologies in the application areas mentioned above; and second, to develop new tools and techniques including new input and output technologies, novel techniques for visualization and techniques for manual and automatic content creation and manipulation.
Transforming the financial system for the public good
Neha Narula, Robleh Ali | E14-244
Staff researchers from the Digital Currency Initiative present early-stage research on digital fiat currency, payments and interoperability, auditing private blockchains, and asset registries.
Urban modeling, simulation, and community engagement
Luis Alonso, Joost Bonsen | E15-368
CityScope is a data-driven, evidence-based decision support tool that can be used in a dynamic, iterative urban design and system integration process. Using a suite of mathematical models and simulation tools, CityScope is designed to facilitate consensus building by allowing participants to explore alternatives and to receive real-time feedback about the impact of design, technology, and policy interventions. In this workshop, we will discuss how these tools are being used to help cities develop innovation districts in Hamburg, Shanghai, Helsinki, and Andorra. Participants will experience CityScope Kendall Square, which is focused on the Volpe Center on a 14-acre parcel adjacent to MIT.
Stacie Slotnick, Media Lab website project lead | E14-514B
Now that we've launched phase one of the Media Lab website, we'd love your feedback and ideas about features that would improve and enhance your Media Lab member experiences, both digital and in-person. No idea is too big, too small, or too crazy! Ideally, participants will have logged into the Media Lab's new site using their insite credentials and then browsed around. We'd like to keep the conversation more about ideation and less about bug reporting (you can send bug reports to firstname.lastname@example.org at any time). Feel free to email Stacie in advance with questions or comments (email@example.com).