This dissertation introduces systems and algorithms designed for efficient and reliable unstructured data analytics, particularly for machine learning applications. It addresses the high costs and unreliability associated with current ML methods by developing systems that provide approximate answers with accuracy guarantees. These systems can be orders of magnitude cheaper than standard approaches and offer statistically valid results for various query types. AI
IMPACT Offers a path to significantly reduce the cost and improve the reliability of ML-driven data analysis for organizations.
RANK_REASON The cluster discusses a dissertation detailing new systems and algorithms for data analytics, which falls under research.
AI-generated summary · Google Gemini · from 3 sources. How we write summaries →