LECTURE 1 TERM 2:
MSIN0097
Predictive Analytics
A P MOORE
DATA-PATTERN-ALGORITHM-MODEL
TYCHO BRAHE 1546 – 1601
A VARIATION OF THE DIKW PYRAMID
DATA
— “A manipulable representation of a past state of (part of) the world”
— Minimum useful set of actions:
— create
— store (persist)
— copy (replicate)
— alter (mutable)
— destroy —
DATA TECHNOLOGIES
8000 BCE 3000 BCE 0 BCE
INFORM ATION PHAISTOS DISK
It is generally agreed that there is not enough context available for a meaningful analysis
Source: https://en.wikipedia.org/wiki/Phaistos_Disc
INFORMATION TECHNOLOGIES
INFORMATION TECHNOLOGIES
mid 1950s
1990s
2010s
Mainframes
Databases
Networks
Servers
Software
Operating Systems Programming languages
SQL
“is it in the database”
Internet Website
Java
TCPIP
SMTP
HTTP (DOM) URLs
HTML (CSS) JSON
Browsing
Blogging
Search – “Google”
“is it in on the web”
Blockchain Consensus algorithm Distributed ledgers Oracles
Digital wallet Transaction blocks
Identities
Titles
Smart contracts Digital ownership
“is it linked to the blockchain”
“is it on chain”
DATA-PATTERN-ALGORITHM-MODEL
PAT TERNS OBSERVATION IS THEORY LADEN
What am I looking for? What is describable? What can I explain? Correlation? Causation?
Utility?
GESTALT PHYCOLOGY
GESTALT PHYCOLOGY
MINDS EYE
The name “Bear” is Homeric, and apparently native to Greece, while the “Wain” tradition is Mesopotamian. Book XVIII of Homer’s Iliad mentions it as “the Bear, which men also call the Wain”.[10] In Latin, these seven stars were known as the “Seven Oxen” (septentriones, from septem triōnēs).[11]Classical Greek mythography identified the “Bear” as the nymph Callisto, changed into a she-bear
by Hera, the jealous wife of Zeus.
Source: Wikipedia
DATA-PATTERN-ALGORITHM-MODEL
MODELS
ALL MODELS ARE WRONG BUT SOME ARE USEFUL
Sun
The Sun is big:
You can fit 1.3m Earths inside the Sun
Source: Jason Major
Ear th
MODELS
MODELS
MODELS
MODELS
MODELS
MODELS
ALL MODELS ARE WRONG THE ROLE OF EXPLANATORY POWER
https://twitter.com/ZonePhysics/status/1156293511790112768
MODELS IN MACHINE LEARNING
MACHINE LEARNING BR ANCHES
We know what the right answer is
We don’t know what the right answer is – but we can recognize a good answer if we find it
We have a way to measure how good our current best answer is, and a method to improve it
Source: Introduction to Reinforcement Learning, David Silver
LECTURE 1 TERM 2:
MSIN0097
Predictive Analytics
A P MOORE