Data Science: Achievements
- Updated
On this page
Exploratory analysis becomes an explicit approach
Established · 1977
Exploratory data analysis emphasizes learning from the structure of observations before imposing a narrow model. Graphs and summaries can expose outliers or assumptions that deserve investigation. They support inquiry; a striking pattern is not automatically a causal explanation.
What is Exploratory Data Analysis? — NIST/SEMATECH statistical handbook · Checked
Large-scale analysis gains a reusable execution model
Demonstrated result · 2004
Dean and Ghemawat’s MapReduce paper described processing data through map and reduce operations while the runtime handled distribution and failures. The reported system showed how analysis could use large clusters without every analyst implementing the coordination machinery. It did not make data quality irrelevant.
MapReduce: Simplified Data Processing on Large Clusters — Dean and Ghemawat / OSDI · Source date: 2004 · Checked
Statistical evidence supports a new astronomical observation
Demonstrated result · Observation: 14 September 2015; paper: 11 February 2016
The LIGO and Virgo collaborations reported the first direct gravitational-wave detection and observation of a binary black-hole merger. Instrumentation, physical modelling and statistical signal analysis worked together. This is a scientific achievement supported by data analysis, not a claim that software alone made the discovery.
Observation of Gravitational Waves from a Binary Black Hole Merger — LIGO Scientific Collaboration and Virgo Collaboration · Source date: 2016-02-11 · Checked