New Data Science Framework for Analysing and Mining Big Data
summary
Proposes a stepwise framework that claims to reduce wasted effort by forcing problem definition, requirement gathering, then data acquisition, fusion, cleansing, feature selection, an 80/10/10 split for modeling and testing, and evaluation tied back to the original goal. Uses satellite plus deployed IoT sensors to build a Big Data platform and a cyber‑physical system for Mekong Delta salinity prediction. Notes most projects are not true big data — nine out of ten projects not big data — and targets students and junior teams.
AI-written summary · about
chapters (7)
- 0:00Start
- 1:21Outline
- 3:15Overview
- 8:59Motivation
- 10:26Methodology
- 13:44Related Work
- 23:57Architecture of the Proposed Framework
