road to driverless cars?
of our planet?
of our planet?
Will “deep learning” put us on the road to driverless cars?
From the readings provided by medical instruments, to ecological sensors monitoring pollution, to digital signals sent from spacecraft circling distant planets, data is the foundation of all scientific exploration. The most basic units of information, data offer scientists a glimpse of truths that are otherwise invisible. In the Information Age, we are witnessing before our eyes a revolution in digital technologies and instruments creating unprecedented amounts of data and the immediate need for new methods to analyze it. Columbia scientists across an astounding array of disciplines are working together to develop innovative, new ways to organize, access, and make use of the vast information we are collectively gathering.
Advances in data science cut across each of our six research themes: neuroscience, climate response, precision medicine, next generation nanoscience, and fundamental science. Success in each of these areas depends on our ability to effectively harness Big Data in creative new ways.
Data Science Institute
The Data Science Institute at Columbia is training the next generation of data scientists and developing innovative technology to serve society. With nearly 200 affiliated faculty working in a wide range of disciplines, including statistics, earth and environmental sciences, psychology, and physics within the Arts & Sciences, the Data Science Institute is fostering collaboration and advancing techniques to gather and interpret data, and to address the urgent problems facing society.
Probability & Society
Understanding and controlling randomness is a unifying goal that connects faculty across Columbia and is crucial to solving global challenges. It provides a framework for studying complex systems such as those in climate, politics, earth science, neuroscience, transportation, economics, and bioinformatics. Columbia is taking the innovative step to create a first-of-its-kind initiative that will bring together faculty, students, and the broader community to work on and understand probability and its application to improving research and society.
“Pixel Approximate Entropy” quantifies visual complexity by providing a score that automatically identifies difficult charts. The technique could help users in emergency settings to read data at a glance and make better decisions faster.
Tamar Mitts, whose research uses big data, machine learning, and text analysis to study conflict, radicalization and violent extremism, is a faculty member at Columbia’s Data Science Institute.
Ten New York City science and math teachers spent the summer learning how to master the most advanced techniques in wireless technology with help from researchers at Columbia University, Columbia’s Data Science Institute, and NYU.