Thu 25 Apr 11:45: Stochastic approximation with heavy tailed noise TMLW02 - SGD: stability, momentum acceleration and heavy tails
TMLW02 - SGD: stability, momentum acceleration and heavy tails
- Speaker: Vivek Borkar (Indian Institute of Technology Bombay)
- Thursday 25 April 2024, 11:45-12:30
- Venue: External.
- Series: Isaac Newton Institute Seminar Series; organiser: nobody.
Thu 25 Apr 11:00: Stochastic Gradient Descent with Adaptive Data TMLW02 - SGD: stability, momentum acceleration and heavy tails
TMLW02 - SGD: stability, momentum acceleration and heavy tails
- Speaker: Jing Dong (Columbia University)
- Thursday 25 April 2024, 11:00-11:45
- Venue: External.
- Series: Isaac Newton Institute Seminar Series; organiser: nobody.
Thu 25 Apr 09:45: Sharp Characterization and Control of Global Dynamics of SGDs with Heavy Tails TMLW02 - SGD: stability, momentum acceleration and heavy tails
TMLW02 - SGD: stability, momentum acceleration and heavy tails
- Speaker: Xingyu Wang (Northwestern University)
- Thursday 25 April 2024, 09:45-10:30
- Venue: External.
- Series: Isaac Newton Institute Seminar Series; organiser: nobody.
Wed 24 Apr 15:15: Aggregating regularly varying vectors: phenomena of a few large jumps TMLW02 - SGD: stability, momentum acceleration and heavy tails
TMLW02 - SGD: stability, momentum acceleration and heavy tails
- Speaker: Bikramjit Das (Singapore University of Technology and Design)
- Wednesday 24 April 2024, 15:15-16:00
- Venue: External.
- Series: Isaac Newton Institute Seminar Series; organiser: nobody.
Wed 24 Apr 14:30: Algorithmic stability for heavy-tailed SGD TMLW02 - SGD: stability, momentum acceleration and heavy tails
TMLW02 - SGD: stability, momentum acceleration and heavy tails
- Speaker: Lingjiong Zhu (Florida State University)
- Wednesday 24 April 2024, 14:30-15:15
- Venue: External.
- Series: Isaac Newton Institute Seminar Series; organiser: nobody.
Wed 24 Apr 11:45: Learning Operators - Recent Advances TMLW02 - SGD: stability, momentum acceleration and heavy tails
TMLW02 - SGD: stability, momentum acceleration and heavy tails
- Speaker: Nicole Mücke (None / Other)
- Wednesday 24 April 2024, 11:45-12:30
- Venue: External.
- Series: Isaac Newton Institute Seminar Series; organiser: nobody.
Wed 24 Apr 11:00: PCF-GAN: generating sequential data via the characteristic function of measures on the path space TMLW02 - SGD: stability, momentum acceleration and heavy tails
TMLW02 - SGD: stability, momentum acceleration and heavy tails
- Speaker: Hao Ni (University College London)
- Wednesday 24 April 2024, 11:00-11:45
- Venue: External.
- Series: Isaac Newton Institute Seminar Series; organiser: nobody.
Wed 24 Apr 09:45: About small jumps of Lévy processes: approximations and estimation TMLW02 - SGD: stability, momentum acceleration and heavy tails
TMLW02 - SGD: stability, momentum acceleration and heavy tails
- Speaker: Ester Mariucci (Université de Versailles Saint-Quentin-en-Yvelines)
- Wednesday 24 April 2024, 09:45-10:30
- Venue: External.
- Series: Isaac Newton Institute Seminar Series; organiser: nobody.
Tue 23 Apr 09:45: Adaptive Importance Sampling for accelerating the minimization of tail risks TMLW02 - SGD: stability, momentum acceleration and heavy tails
TMLW02 - SGD: stability, momentum acceleration and heavy tails
- Speaker: Karthyek Rajhaa Annaswamy Murthy (Singapore University of Technology and Design)
- Tuesday 23 April 2024, 09:45-10:30
- Venue: External.
- Series: Isaac Newton Institute Seminar Series; organiser: nobody.
Mon 22 Apr 16:45: Reception TMLW02 - SGD: stability, momentum acceleration and heavy tails
TMLW02 - SGD: stability, momentum acceleration and heavy tails
- Speaker:
- Monday 22 April 2024, 16:45-19:00
- Venue: External.
- Series: Isaac Newton Institute Seminar Series; organiser: nobody.
Mon 22 Apr 16:00: Title TBA TMLW02 - SGD: stability, momentum acceleration and heavy tails
TMLW02 - SGD: stability, momentum acceleration and heavy tails
- Speaker: Jose Blanchet (Stanford University)
- Monday 22 April 2024, 16:00-16:45
- Venue: External.
- Series: Isaac Newton Institute Seminar Series; organiser: nobody.
Mon 22 Apr 14:30: Progress Report Presentations TMLW02 - SGD: stability, momentum acceleration and heavy tails
TMLW02 - SGD: stability, momentum acceleration and heavy tails
- Speaker:
- Monday 22 April 2024, 14:30-16:00
- Venue: External.
- Series: Isaac Newton Institute Seminar Series; organiser: nobody.
Mon 22 Apr 11:45: Heavy-tailed compound renewal and L\'evy processes with negative drift: maxima over random time intervals TMLW02 - SGD: stability, momentum acceleration and heavy tails
TMLW02 - SGD: stability, momentum acceleration and heavy tails
- Speaker: Sergey Foss (Heriot-Watt University)
- Monday 22 April 2024, 11:45-12:30
- Venue: External.
- Series: Isaac Newton Institute Seminar Series; organiser: nobody.
Mon 22 Apr 09:45: Exit Times and Extremes of Fractional Brownian motion and Spectrally Negative Levy Processes TMLW02 - SGD: stability, momentum acceleration and heavy tails
TMLW02 - SGD: stability, momentum acceleration and heavy tails
- Speaker: Ceren Vardar Acar (Middle East Technical University)
- Monday 22 April 2024, 09:45-10:30
- Venue: External.
- Series: Isaac Newton Institute Seminar Series; organiser: nobody.
Mon 10 Jun 14:50: TBA ADIW04 - Anti-Diffusion in Multiphase and Active Flows
ADIW04 - Anti-Diffusion in Multiphase and Active Flows
- Speaker: Joshua Robinson (University of Bristol)
- Monday 10 June 2024, 14:50-15:20
- Venue: Seminar Room 1, Newton Institute.
- Series: Isaac Newton Institute Seminar Series; organiser: nobody.
Tue 16 Apr 14:00: Lie Fu
- Speaker:
- Tuesday 16 April 2024, 14:00-15:00
- Venue: Telecommunications Room.
- Series: Isaac Newton Institute Seminar Series; organiser: nobody.
Fri 26 Apr 15:30: Leveraging multi-study, multi-outcome data to improve external validity and efficiency of clinical trials for managing schizophrenia
As data sources have become more plentiful and readily accessible, the practice of data fusion has become increasingly ubiquitous. However, when the focus is on a causal effect on a particular outcome, a major limitation is that this outcome may not be available in all data sources. In fact, different randomized experiments or observational studies of a common exposure will often focus on potentially related, yet distinct outcomes. One such example is the Database of Cognitive Training and Remediation Studies (DoCTRS), which consists of several randomized trials of the effect of cognitive remediation therapy on various outcomes among patients with schizophrenia. We develop causally principled methodology for fusing data sets when multiple outcomes are observed across studies, which leverages outcomes of secondary interest as informative proxies for the missing outcome of primary interest, thereby maximizing power and efficiency by making full use of the available data. As this methodology relies on a key transportability assumption, we also develop methods to assess the degree of sensitivity to violations of this assumption. We apply this methodology to data from the DoCTRS trials to make improved causal inferences about the effectiveness of cognitive remediation therapy on cognition among patients with schizophrenia.
- Speaker: Caleb Miles (Columbia University)
- Friday 26 April 2024, 15:30-17:00
- Venue: MR12, Centre for Mathematical Sciences, Wilberforce Road, Cambridge.
- Series: Causal Inference Reading Group; organiser: Qingyuan Zhao.
Fri 26 Apr 15:30: Leveraging multi-study, multi-outcome data to improve external validity and efficiency of clinical trials for managing schizophrenia
As data sources have become more plentiful and readily accessible, the practice of data fusion has become increasingly ubiquitous. However, when the focus is on a causal effect on a particular outcome, a major limitation is that this outcome may not be available in all data sources. In fact, different randomized experiments or observational studies of a common exposure will often focus on potentially related, yet distinct outcomes. One such example is the Database of Cognitive Training and Remediation Studies (DoCTRS), which consists of several randomized trials of the effect of cognitive remediation therapy on various outcomes among patients with schizophrenia. We develop causally principled methodology for fusing data sets when multiple outcomes are observed across studies, which leverages outcomes of secondary interest as informative proxies for the missing outcome of primary interest, thereby maximizing power and efficiency by making full use of the available data. As this methodology relies on a key transportability assumption, we also develop methods to assess the degree of sensitivity to violations of this assumption. We apply this methodology to data from the DoCTRS trials to make improved causal inferences about the effectiveness of cognitive remediation therapy on cognition among patients with schizophrenia.
- Speaker: Caleb Miles (Columbia University)
- Friday 26 April 2024, 15:30-17:00
- Venue: MR12, Centre for Mathematical Sciences, Wilberforce Road, Cambridge.
- Series: Causal Inference Reading Group; organiser: Qingyuan Zhao.
Fri 26 Apr 15:45: Final Remarks and Close OFBW65 - Connecting Heavy Tails and Differential Privacy in Machine Learning
OFBW65 - Connecting Heavy Tails and Differential Privacy in Machine Learning
- Speaker: Jorge Gonzalez-Cazares (National Autonomous University of Mexico (UNAM), IIMAS)
- Friday 26 April 2024, 15:45-16:00
- Venue: External.
- Series: Isaac Newton Institute Seminar Series; organiser: nobody.
Fri 26 Apr 10:30: Welcome and Introduction OFBW65 - Connecting Heavy Tails and Differential Privacy in Machine Learning
OFBW65 - Connecting Heavy Tails and Differential Privacy in Machine Learning
- Speaker: Jorge Gonzalez-Cazares (National Autonomous University of Mexico (UNAM), IIMAS), Sofia Sanz Del Pino (Isaac Newton Institute)
- Friday 26 April 2024, 10:30-10:40
- Venue: External.
- Series: Isaac Newton Institute Seminar Series; organiser: nobody.