Business analytics
The golden spiral
Look into a sunflower and its seeds line up in spirals, usually a Fibonacci number of them. Usually: in a citizen-science study published in 2016, volunteers counted the spirals on 657 sunflowers, and roughly a quarter of the counts broke the rule. Analytics is finding the pattern, then counting the exceptions honestly.
Swinton, Ochu & MSI Turing’s Sunflower Consortium, Royal Society Open Science, 2016
Bring a table, or start with an example. Explore it first, then predict an outcome, find structure without labels, or forecast a seasonal series. A table loaded under Explore is available on the other pages, and nothing leaves your browser.
Explore your data
Bring a table: a quality check, descriptive statistics, correlations, and regression with its diagnostics.
- Data-quality audit
- Describe
- Correlate
- Regress (with HC3)
Supervised learning
Predict a yes-or-no outcome, comparing logistic regression, a decision tree, nearest neighbours and a random forest fairly.
- Model comparison by cross-validation
- Decision tree you can read
- Feature importance
- Logistic regression in depth
Unsupervised learning
Find structure without labels: groups of similar rows and the few directions that explain most of the variation.
- k-means
- Hierarchical clustering
- Principal components
Forecasting
Carry a seasonal series forward with a range, after checking the method on a season it never saw.
- Holt-Winters with 95% intervals
- Held-back-season check
Sources
- Swinton, J., Ochu, E. & MSI Turing’s Sunflower Consortium (2016). Novel Fibonacci and non-Fibonacci structure in the sunflower: results of a citizen science experiment. Royal Society Open Science 3(5). doi:10.1098/rsos.160091. Spiral-count shares as summarised by Plus Magazine.