COMP9417 Aims
09s1: Machine Learning and Data Mining This lecture will provide the basis for you to be able to describe the motivation, scope and some application areas of machine learning. Following it you should be able to: describe the general learning problem March 12, 2008 state some of the steps in setting up a learning problem list some applications of machine learning Acknowledgement: Material derived from slides for the book Machine Learning, Tom Mitchell, McGraw-Hill, 1997 http://www-2.cs.cmu.edu/~tom/mlbook.html list some issues in machine learning
Introduction to Machine Learning
COMP9417: March 11, 2009
Introduction to Machine Learning: Slide 1
Overview
[Recommended reading: Mitchell, Chapter 1] [Recommended exercises: 1.1,1.2, optionally 1.5] Why Machine Learning? What is a well-dened learning problem? An example: learning to play checkers (draughts) What questions should we ask about Machine Learning?
Why Machine Learning
Considerable progress in algorithms and theory Growing ood of online data Increasing computational power Many successful commercial/scientic applications
COMP9417: March 11, 2009
Introduction to Machine Learning: Slide 2
COMP9417: March 11, 2009
Introduction to Machine Learning: Slide 3
Three niches for machine learning: Data mining: using historical data to improve decisions medical records medical knowledge Software applications we cant program by hand autonomous robots speech recognition Self customizing programs Web sites that learn user interests
Some denitions
machine learning the science of algorithmic methods of learning from experience with the goal of improving performance on selected tasks data mining the use of machine learning or statistical algorithms to search large amounts of data for hidden patterns or relationships that are interesting and potentially useful
COMP9417: March 11, 2009
Introduction to Machine Learning: Slide 4
COMP9417: March 11, 2009
Introduction to Machine Learning: Slide 5
Typical data mining Task
Patient103 time=1
Age: 23 FirstPregnancy: no Anemia: no Diabetes: no PreviousPrematureBirth: no Ultrasound: ? Elective CSection: ? Emergency CSection: ? ...
Datamining Result
Patient103 time=1 Patient103 time=2
Age: 23 FirstPregnancy: no Anemia: no Diabetes: YES PreviousPrematureBirth: no Ultrasound: abnormal Elective CSection: no Emergency CSection: ? ...
Patient103 time=2
Age: 23 FirstPregnancy: no Anemia: no Diabetes: YES PreviousPrematureBirth: no Ultrasound: abnormal Elective CSection: no Emergency CSection: ? ...
...
Patient103 time=n
Age: 23 FirstPregnancy: no Anemia: no Diabetes: no PreviousPrematureBirth: no Ultrasound: ? Elective CSection: no Age: 23 FirstPregnancy: no Anemia: no Diabetes: no PreviousPrematureBirth: no Ultrasound: ? Elective CSection: ? Emergency CSection: ? ...
...
Patient103 time=n
Age: 23 FirstPregnancy: no Anemia: no Diabetes: no PreviousPrematureBirth: no Ultrasound: ? Elective CSection: no
Emergency CSection: Yes
...
Emergency CSection: Yes
...
Given: 9714 patient records, each describing a pregnancy and birth Each patient record contains 215 features Learn to predict: Classes of future patients at high risk for Emergency Cesarean Section
COMP9417: March 11, 2009 Introduction to Machine Learning: Slide 6
One of 18 learned rules: If No previous vaginal delivery, and Abnormal 2nd Trimester Ultrasound, and Malpresentation at admission Then Probability of Emergency C-Section is 0.6 Over training data: 26/41 = .63, Over test data: 12/20 = .60
COMP9417: March 11, 2009 Introduction to Machine Learning: Slide 7
Credit Risk Analysis
Customer103: (time=t0)
Years of credit: 9 Loan balance: $2,400 Income: $52k Own House: Yes Other delinquent accts: 2 Max billing cycles late: 3 Profitable customer?: ? ...
Other Prediction Problems
Customer purchase behavior:
Customer103: (time=t0)
Sex: M Age: 53 Income: $50k Own House: Yes MS Products: Word Computer: 386 PC Purchase Excel?: ? ... Years of credit: 9 Loan balance: $4,500 Income: ? Own House: Yes Other delinquent accts: 3 Max billing cycles late: 6
Customer103: (time=t1)
Years of credit: 9 Loan balance: $3,250 Income: ? Own House: Yes Other delinquent accts: 2 Max billing cycles late: 4 Profitable customer?: ? ...
...
Customer103: (time=tn)
Customer103: (time=t1)
Sex: M Age: 53 Income: $50k Own House: Yes MS Products: Word Computer: Pentium Purchase Excel?: ? ...
...
Customer103: (time=tn)
Sex: M Age: 53 Income: $50k Own House: Yes MS Products: Word Computer: Pentium
Profitable customer?: No
...
Purchase Excel?: Yes
...
Rules learned from synthesized data: If Other-Delinquent-Accounts > 2, and Number-Delinquent-Billing-Cycles > 1 Then Profitable-Customer? = No [Deny Credit Card application] If Other-Delinquent-Accounts = 0, and (Income > $30k) OR (Years-of-Credit > 3) Then Profitable-Customer? = Yes [Accept Credit Card application]
COMP9417: March 11, 2009 Introduction to Machine Learning: Slide 8
Customer retention:
Customer103: (time=t0)
Sex: M Age: 53 Income: $50k Own House: Yes Checking: $5k Savings: $15k Currentcustomer?: yes ...
Customer103: (time=t1)
Sex: M Age: 53 Income: $50k Own House: Yes Checking: $20k Savings: $0 Currentcustomer?: yes ...
...
Customer103: (time=tn)
Sex: M Age: 53 Income: $50k Own House: Yes Checking: $0 Savings: $0
Currentcustomer?: No
COMP9417: March 11, 2009
Introduction to Machine Learning: Slide 9
Process optimization:
Product72:
Stage: mix Mixingspeed: 60rpm Viscosity: 1.3 Fat content: 15% Density: 2.8 Spectral peak: 2800 Product underweight?: ?? ... (time=t0)
Tasmanian Apple Thinning
Product72:
Stage: cook Temperature: 325 Viscosity: 3.2 Fat content: 12% Density: 1.1 Spectral peak: 3200 Product underweight?: ?? ... (time=t1)
...
Product72:
Stage: cool
(time=tn)
Fanspeed: medium Viscosity: 1.3 Fat content: 12% Density: 1.2 Spectral peak: 3100
Product underweight?: Yes
...
Apple orchards are important in primary production in Tasmania, and there has been a long history in the process of apple thinning. Apples are naturally biennial bearing,. Trees ower heavily one year producing a large crop of small fruit (called the On year) followed by light owering the next year with a small crop of large poor quality fruit. Thinning is most economically done by applying sprays of chemicals that act similarly to plant hormones and cause the abortion of owers and fruitlets at an early stage of development. Early thinning favours the development of the desirable high density of cells in the fruit. Orchardists decision about concentration of thinning agent at blossom time. If concentration too low, then thinning is not eective and cost of hand thinning is prohibitive,. If the concentration too high, then risk of losing all the fruit. Decision is dicult because of large number of variables to be taken into account.
COMP9417: March 11, 2009
Introduction to Machine Learning: Slide 10
COMP9417: March 11, 2009
Introduction to Machine Learning: Slide 11
trees - cultivar, rootstock and age. physiology - previous crop, vigour, number of blossom buds. pruning - severity of detailed pruning, limb thinning, and penetration of light into the canopy. market - size of fruit required for the market. spraying - type of spray machinery and volume of water to be used in the machinery. 60 tasks, (some with 50 decision tree leaves (i.e. rule paths), plus 30 other variables and 40 procedures supported by a customized help le of 5,000 words.
BG Gas Drilling - Stuck Pipe
Drilling is a hugely expensive process, with daily costs for a North Sea operation typically incurring rig costs of around $50,000 per day. Clearly, anything that helps to reduce the time when a drilling rig is not productive has the potential to achieve huge savings.
COMP9417: March 11, 2009 Introduction to Machine Learning: Slide 12 COMP9417: March 11, 2009 Introduction to Machine Learning: Slide 13
Daily report data from two databases: One of which was old and included incomplete or absent data - particularly IADC (International Association of Drilling Contractors) codes. The other database was compiled more recently and included a large amount of additional data about well site geology, drilling costs, etc. Sixty recorded occurrences of Stuck Pipe in 170 BG wells. Possible to mine the data and to determine trends. Much of the time invested by the project team has concentrated on getting data in good order. Results indicate that length of time the hole has been open; the properties of the drilling mud; and the frequency with which the mud is conditioned all play a signicant role in the incidence of Stuck Pipe.
Nissan - Car selection
Starting from the basic choices of 3 alternative engines, 3 types of suspension, 2 types of transmission, 9 colours and 3 styles of seat fabric, customers can go far further and create a car to suit their own personality.
COMP9417: March 11, 2009 Introduction to Machine Learning: Slide 14 COMP9417: March 11, 2009 Introduction to Machine Learning: Slide 15
With 670,000 possible combinations, it is a totally new concept says Takao Ohmura, Sales Manager of Tokyo Nissan Computer Systems. A guidebook explains the options in table form and we were able to input these tables into XpertRule. Normally it is dicult to utilise such a large matrix, but XpertRule was able to automatically generate a decision tree structure to arrive at the correct model, from attributes and values in the tables. It met our three major requirements: (1) the model selection and check must be completed in minutes: three (2) the ability to run on Nissan dealers hardware, and (3) ease of maintaining the system after the launch of the Cero model.
Channel 4 TV scheduling
During the day, Channel 4s strength is the housewife market whilst in the evenings Channel 4s strength lies in its varied targeting ability. In comparison with ITV, Channel 4 audiences contain a greater proportion of younger, lighter, up-market, male viewers (audience research has also identied Channel 4s ability to target cluster groups dened by names such as Progressive Priscillas and Free-thinking Franks). Advertisers may specify to have commercials placed rst in the break, last in the break or Top & Tail in a break making break sequencing a challenge if optimal use of airtime is to be achieved. Denition of a knowledge-based system to solve the problem requires observation of a number of prioritised rules: Top of the list is the need for no overlaps or gaps, with Top and Tail or First and Last network spots also receiving high priority. Lower down the list are First and Last Super-macro spots and non-reporting Super-macros sequenced to play at the same time.
COMP9417: March 11, 2009 Introduction to Machine Learning: Slide 17
COMP9417: March 11, 2009
Introduction to Machine Learning: Slide 16
Optimization problems as the number of possible combinations grows it becomes impractical to try all combinations to arrive at a solution in a reasonable time. Rule of thumb can be used to narrow down options but, in most cases, good rules are not available or are dicult to capture. Numerical optimization techniques are currently available in most advanced spreadsheets, but these tend to be incapable of optimizing problems involving sequencing or scheduling and they are exploitation rather than exploration techniques. The solution involved the use of genetic algorithm techniques which allows the exploration of large search spaces for optimal or near optimal solutions.
Problems Too Dicult to Program by Hand
ALVINN [Pomerleau] drives 70 mph on highways !
COMP9417: March 11, 2009
Introduction to Machine Learning: Slide 18
COMP9417: March 11, 2009
Introduction to Machine Learning: Slide 19
Sharp Left
Straight Ahead
Sharp Right
Stanley - DARPA Grand Challenge Champion 2005
30 Output Units
4 Hidden Units
30x32 Sensor Input Retina
won 2 million dollars (US), rst team to complete 132 mile course modied VW Touareg R5 with drive-by-wire, took 6 hours 54 minutes averaging over 19 mph seven Pentium M computers, GPS and various sensors localization, mapping and collision avoidance
COMP9417: March 11, 2009
Introduction to Machine Learning: Slide 20
COMP9417: March 11, 2009
Introduction to Machine Learning: Slide 21
Software that adapts to User
Measuring Neural Activity
Brin & Page - PhD students in data mining at Stanford PageRank algorithm (1998) Google business model - technology targets advertisements to users
Botros, van Dijk & Killian (2007) - Cochlear implant adjustment Expert system uses neural response telemetry (ECAP) Decision tree learning - Quinlans C5 and Cubist
COMP9417: March 11, 2009
Introduction to Machine Learning: Slide 22
COMP9417: March 11, 2009
Introduction to Machine Learning: Slide 23
concentration of yeast in the wells of the microtitre trays using the adjacent plate reader and returns the results to the LIMS (although microtitre trays are still moved in and out of incubators manually). Scientic Discovery
The Robot Scientist project (2004)
Figure 1 The Robot Scientist hypothesis-generation and experimentation loop.
248
COMP9417: March 11, 2009 Introduction to Machine Learning: Slide 24
biological entities. The original bioinformatic information for the AAA model was Scientic Discovery taken mainly from the KEGG13 catalogue of metabolism. The model was then tested with all possible auxotrophic experiments involving Robot scientist in the lab a single replacement metabolite, and was altered manually to t the empirical results. To ensure that the model was not over-tted, we carried out all possible auxotrophic experiments with pairs of metabolites. The model correctly predicted at least 98.5% of the experiments (Supplementary Information). To the best of our knowledge, no bioinformatic model has been as thoroughly tested with knockout mutants. Machine learning is the branch of articial intelligence that seeks to develop computer systems that improve their performance automatically with experience14,15. It has much in common with statistics, but differs in having a greater emphasis on algorithms, data representation and making acquired knowledge explicit. The
COMP9417: March 11, 2009
NATURE | VOL 427 | 15 JANUARY 2004 | www.nature.com/nature
Introduction to Machine Learning: Slide 25
Where Is this Headed ?
Mature algorithms decision trees, regression, neural nets, Bayesian methods ... can be applied to standard database relations or at les established software and services industry
Where Is this Headed ?
Opportunity for tomorrow: enormous impact Learn across full mixed-media data Learn across multiple internal databases, plus the web and newsfeeds Learn by active experimentation Learn more complex functions Learn by analogy Cumulative, lifelong learning and adaptation Programming languages and systems with learning embedded ?
COMP9417: March 11, 2009
Introduction to Machine Learning: Slide 26
COMP9417: March 11, 2009
Introduction to Machine Learning: Slide 27
Relevant Disciplines
Articial intelligence Computational complexity theory Statistics Information theory Bayesian methods Control theory Philosophy Psychology and neurobiology Physics ...
COMP9417: March 11, 2009 Introduction to Machine Learn...