Search:
Match:
6 results
Research#Machine Learning📝 BlogAnalyzed: Dec 29, 2025 07:45

Optimization, Machine Learning and Intelligent Experimentation with Michael McCourt - #545

Published:Dec 16, 2021 17:49
1 min read
Practical AI

Analysis

This article summarizes a podcast episode featuring Michael McCourt, Head of Engineering at SigOpt. The discussion centers on optimization, machine learning, and their intersection. Key topics include the technical distinctions between ML and optimization, practical applications, the path to increased complexity for practitioners, and the relationship between optimization and active learning. The episode also delves into the research frontier, challenges, and open questions in optimization, including its presence at the NeurIPS conference and the growing interdisciplinary collaboration between the machine learning community and fields like natural sciences. The article provides a concise overview of the podcast's content.
Reference

“The article doesn't contain a direct quote.”

PermalinkPractical AI
Research#Predictive Maintenance📝 BlogAnalyzed: Dec 29, 2025 07:46

Predictive Maintenance Using Deep Learning and Reliability Engineering with Shayan Mortazavi - #540

Published:Nov 29, 2021 18:58
1 min read
Practical AI

Analysis

This article summarizes a podcast episode featuring Shayan Mortazavi, a data science manager at Accenture. The episode focuses on Mortazavi's presentation at the SigOpt HPC & AI Summit, which detailed a novel deep learning approach for predictive maintenance in oil and gas plants. The discussion covers the evolution of reliability engineering, the use of a residual-based approach for anomaly detection, challenges with LSTMs, and the human labeling requirements for model building. The article highlights the practical application of AI in industrial settings, specifically for preventing equipment failure and damage.
Reference

“In the talk, Shayan proposes a novel deep learning-based approach for prognosis prediction of oil and gas plant equipment in an effort to prevent critical damage or failure.”

PermalinkPractical AI
Research#AI Algorithms📝 BlogAnalyzed: Dec 29, 2025 07:49

Constraint Active Search for Human-in-the-Loop Optimization with Gustavo Malkomes - #505

Published:Jul 29, 2021 18:19
1 min read
Practical AI

Analysis

This article from Practical AI discusses a new algorithmic solution for iterative model search, focusing on constraint active search. The guest, Gustavo Malkomes, a research engineer at Intel (via SigOpt), explains his paper on multi-objective experimental design. The algorithm allows teams to identify parameter configurations that satisfy constraints in the metric space, rather than optimizing specific metrics. This approach enables efficient exploration of multiple metrics simultaneously, making it suitable for real-world, human-in-the-loop scenarios. The article highlights the potential of this method for informed and intelligent experimentation.
Reference

“This new algorithm empowers teams to run experiments where they are not optimizing particular metrics but instead identifying parameter configurations that satisfy constraints in the metric space.”

PermalinkPractical AI
Research#llm📝 BlogAnalyzed: Dec 29, 2025 08:08

Automated Model Tuning with SigOpt - #324

Published:Dec 9, 2019 20:43
1 min read
Practical AI

Analysis

This article summarizes a TWIML Democast episode featuring SigOpt's Co-Founder and CEO, Scott Clark. The focus is on the SigOpt platform and its capabilities for automated model tuning. The article highlights a live demo, suggesting a practical, hands-on approach to understanding the platform. The primary takeaway is the introduction of SigOpt and its function in optimizing machine learning models. The article directs readers to a video demo for a more comprehensive understanding.

Key Takeaways

Reference

“This episode is best consumed by watching the corresponding video demo, which you can find at twimlai.com/talk/324.”

PermalinkPractical AI
Research#AI Platforms📝 BlogAnalyzed: Dec 29, 2025 08:13

Supporting Rapid Model Development at Two Sigma with Matt Adereth & Scott Clark - TWIML Talk #273

Published:Jun 11, 2019 17:16
1 min read
Practical AI

Analysis

This article summarizes a podcast episode from Practical AI featuring Matt Adereth from Two Sigma and Scott Clark from SigOpt. The discussion centers around Two Sigma's modeling platform, its users, and the challenges encountered in production and modeling. The conversation also explores Two Sigma's approach to experimentation and the motivations for companies to invest in platforms, optimization, and automation. The focus is on practical applications and insights into the development and deployment of AI models within a financial context, highlighting the importance of platforms and automation for efficiency.
Reference

“The article doesn't contain a direct quote, but rather outlines the topics discussed.”

PermalinkPractical AI
Research#AI Optimization📝 BlogAnalyzed: Dec 29, 2025 08:38

Bayesian Optimization for Hyperparameter Tuning with Scott Clark - TWiML Talk #50

Published:Oct 2, 2017 21:58
1 min read
Practical AI

Analysis

This article summarizes a podcast episode featuring Scott Clark, CEO of Sigopt, discussing Bayesian optimization for hyperparameter tuning. The conversation delves into the technical aspects of this process, including exploration vs. exploitation, Bayesian regression, heterogeneous configuration models, and covariance kernels. The article highlights the depth of the discussion, suggesting it's geared towards a technically inclined audience. The focus is on the practical application of Bayesian optimization in model parameter tuning, a crucial aspect of AI development.
Reference

“We dive pretty deeply into that process through the course of this discussion, while hitting on topics like Exploration vs Exploitation, Bayesian Regression, Heterogeneous Configuration Models and Covariance Kernels.”

PermalinkPractical AI