ML - Regression

Abstract

Using Linear Regression to Predict Numerical Data

Introduction: Mixing Colours

Interpolation:

• between TWO colours, both colours inclusive using a straight line between them
• between several different colours?
  • by mixing “equal proportions” of each
  • Proportions based on “distance” from each colour
  • On a “plane” with these points

Some Examples from Drama

1. Legally Blonde:

2. Greek Chorus Explained:

3. Sutradhar in Indian Drama

From Encyclopaedia Brittanica

Classical Indian drama had as its elements poetry, music, and dance, with the sound of the words assuming more importance than the action or the narrative; therefore, staging was basically the enactment of poetry. The reason that the productions, in which scenes apparently follow an arbitrary order, seem formless to Westerners is that playwrights use much simile and metaphor. Because of the importance of the poetic line, a significant character is the storyteller or narrator, who is still found in most Asian drama. In Sanskrit drama the narrator was the sūtra-dhāra, “the string holder,” who set the scene and interpreted the actors’ moods. Another function was performed by the narrator in regions in which the aristocratic vocabulary and syntax used by the main characters, the gods and the nobles, was not understood by the majority of the audience. The narrator operated first through the use of pantomime and later through comedy.

Discussion

• Interpolation
• Extrapolation
• Calculating the optimum values for m and c, given x and y, for $y=mx+c$

Playing with Orange: Paint My Data

Regression Plane

Interactive Regression Plane

Workflow in Orange

Let us “draw inspiration” from the picture above, and see if we can replicate it. We will fire up Orange, paint some data and see if we can fit a linear regression ML model to it.

Here is the Orange file for you to download. Open this file in Orange.

Workflow in Radiant

Workflow in R

Conclusion