Video: DAX function Related() + RelatedTable()
söndag 14 augusti 2016
Live Data Stream --> Data Lake --> PowerBI
Pic. PowerBI-monitor
Pic. PubNub
Pic. link
Pic.: Get a Make a POST request,
Pic. Updating SQL-server via PowerQuery
fredag 12 augusti 2016
Sales & Budget + Dimensions, Predictions + etc
Useful
Pic. Sales & Budget
Pic. Useful
-----
Co-operative rules/formulas/procedures: each person is an object with several attributes.
The co-operative procedures are 'context specific'.
Example two men in a rowing boat can co-operate to row in the same direction. The contextual target/objective is to get ashore. The limiting constraints may be that one have a broken arm. So with least input, considering the constraints, get most output. The limiting constraints includes satisfaction based on dissatisfaction (or utility* based on dis-utility). Happiness of the many does not out-weight the sorrows of the few. The right behaviours from rewards/motivations;things that moves us.
* Value dimension can be more/less.
Source: link
Social business --> 'human entities' attributes
Choose a granularity that is understandable, and actionable
a) Missing, Errors, Outliers, ? Repair or Disregard ?
b) Too many categories, change ordinal categories to values, avoid collinearity issues,
Source: link
The curse of dimensionality
Even millions images are not really big in the context of the curse of dimensionality.
The Predictive power reduces as the Dimensionality increases, known as the Hughes effect.
"Even in the simplest case of d binary variables, the number of possible combinations already is O(2d), exponential in the dimensionality. Naively, each additional dimension doubles the effort needed to try all combinations."
---
Read: Predictive modeling,
Pic. Interesting
Pic. Tableau article (IFTTT)
Pic. ggfortify
Pic. Sales & Budget
Pic. Useful
-----
Co-operative rules/formulas/procedures: each person is an object with several attributes.
The co-operative procedures are 'context specific'.
Example two men in a rowing boat can co-operate to row in the same direction. The contextual target/objective is to get ashore. The limiting constraints may be that one have a broken arm. So with least input, considering the constraints, get most output. The limiting constraints includes satisfaction based on dissatisfaction (or utility* based on dis-utility). Happiness of the many does not out-weight the sorrows of the few. The right behaviours from rewards/motivations;things that moves us.
* Value dimension can be more/less.
Source: link
Social business --> 'human entities' attributes
Choose a granularity that is understandable, and actionable
- When selecting predictor variables, keep in mind that you want to gather a maximum amount of information from a minimum number of variables to avoid the curse of dimensionality without overfitting or underfitting.
a) Missing, Errors, Outliers, ? Repair or Disregard ?
b) Too many categories, change ordinal categories to values, avoid collinearity issues,
Source: link
The curse of dimensionality
Even millions images are not really big in the context of the curse of dimensionality.
The Predictive power reduces as the Dimensionality increases, known as the Hughes effect.
"Even in the simplest case of d binary variables, the number of possible combinations already is O(2d), exponential in the dimensionality. Naively, each additional dimension doubles the effort needed to try all combinations."
Source: link
---
Read: Predictive modeling,
Pic. Interesting
Pic. Tableau article (IFTTT)
Pic. ggfortify
Prenumerera på:
Inlägg (Atom)