From Visual Analytics and Applications
Jump to navigation
Jump to search
VAST Challenge 2018:Like a duck to water
Methodology & Design
Methodology
| Description |
Illustration |
| 1.Selecting the right chemicals
Pollution in the waters of the reserve is defined by many measures but the data set only provides enough data for a set of measures
|
|
| 2.Visualizing the overall trend and specific changes
The chemical measures were visualized per quarter and year vs Location. This helps understand the trend for each selected chemical
|
|
| 3. Visual interpretation and distribution analysis
For better user experience I have elaborated a dashboard with a visual representation of each chemical in each location with its specific distribution in a box plot.
|
|
| 4.Outliers analysis
With a control chart the user can determine easily specific outliers for each chemical selecting 1, 2 or 3 standard deviations from the average.Upper bound: WINDOW_AVG(SUM([Value])) + ( WINDOW_STDEV( SUM([Value])) * [Standard Deviations] ), Lower bound: WINDOW_AVG(SUM([Value])) - ( WINDOW_STDEV( SUM([Value])) * [Standard Deviations] ). I have also created a measure of Outliers with the following formula: SUM([Value]) < [Lower Bound] OR SUM([Value]) > [Upper Bound]
|
|
| 5.Analysis of anomalies and sampling recommendations
To understand the sampling strategy in the dataset and suggest improvements I have created a dashboard
|
|
| 6.Detailed distribution of samples
I have designed a dashboard to help the user understand the anomalies in the sampling data
|
|
VAST Challenge 2018:Like a duck to water
