Tommi Jaakkola, professor of computer science and engineering at MIT, has been named the first Thomas Siebel at the Department of Electrical Engineering and Computer Science (EECS) and the Institute for Data, Systems and Society (IDSS).
The nominations were announced by Anantha Chandrakasan, head of EECS and the Vannevar Bush Professor at EECS, and Munther A. Dahleh, director of IDSS and William Coolidge Professor at EECS. “This nomination recognizes Professor Jaakkola’s leadership in the field of machine learning and his outstanding mentorship and educational contributions,” wrote Chandrakasan and Dahleh in a message to EECS faculty. “Professor Jaakkola is internationally renowned in the fields of machine learning and natural language processing, as well as computational biology. He is widely respected as an original researcher and has made significant contributions.”
The up-to-date professorship was created thanks to the generous contribution of veteran software entrepreneur Thomas Siebel, chairman and CEO of C3IoT. Siebel is well known at MIT for establishing the Siebel Scholars program, which annually provides support to 16 MIT graduate students (five in EECS, five in the Department of Biological Engineering, five in the MIT Sloan School of Management, and one majoring in energy science ).
At the core of Jaakkola’s research are inferential and estimation questions in convoluted modeling tasks, from developing the underlying theory and related algorithms to translating such advances into applications. He has been a leading contributor to the development of distributed probabilistic inference algorithms from the field’s inception to its current status as a well-established field of research.
From a modeling perspective, Jaakkola’s work covers a wide range of areas, from the intersection of generative and discriminative modeling, rethinking modeling from the perspective of randomization and combinatorial optimization, to retrieval issues related to continuous object embedding. In natural language processing (NLP), its contributions to solving tough combinatorial inference problems such as parsing natural language, developing deep convolutional representations of text, and reformulating convoluted models to reveal interpretable rationales for predictions. Several of his works have received best publication awards at leading events.
Additionally, Jaakkola “made outstanding contributions to education,” noted Chandrakasan and Dahleh. He founded and oversaw the development of a master’s course in machine learning, teaching it for many years until Professor Leslie Kaelbling took it over for further development. With Professor Regina Barzilay, he developed an undergraduate machine learning course that currently enrolls over 500 students per semester. He modernized the advanced NLP course, again taught by Barzilay, from the point of view of a neural approach to NLP. In 2015, Jaakkola received the Jamieson Award for Excellence in Teaching in recognition of his contributions to education.
He has also made valuable professional contributions in his field and within EECS. He has held editorial positions in prestigious magazines such as “The” and “. He has also co-chaired or supervised areas of major conferences, including the Conference on Neural Information Processing Systems (NIPS), the Conference on Uncertainty in Artificial Intelligence (UAI), and the Conference on Artificial Intelligence and Statistics (AISTATS). He was a member of the EECS Faculty Search Committee for many years and also served on other committees. He also contributed to the career paths of many students and graduate students he supervised and mentored at MIT. Former students and postdocs from his research group currently hold positions at leading universities such as MIT, Carnegie Mellon University, and the University of California at Berkeley.
As an associate member of IDSS, Jaakkola played a key role in both the hiring and recruitment of statistical researchers and the establishment of statistical programs. From the beginning, he served on the IDSS statistics research fellow search committee and worked with the IDSS statistics PhD committee to develop an application for a dual PhD in statistics. He is also a participant in the MicroMasters program dedicated to statistics and data science.