en
0.25
0.5
0.75
1.25
1.5
1.75
2
Datamining "Looking backward, looking forward"
Published on Jan 16, 20133219 Views
The breakthrough of data mining in the mid-nineties can be explained as the culmination of several independent developments: the convergence of machine learning research, the maturity of data base tec
Related categories
Chapter list
Data Mining, Looking backward, looking forward00:00
Data Mining00:23
Looking backward 1985-2012 More than 200 AI,ML,DM projects01:05
Looking forward04:41
1992: Captains04:55
Plan board06:48
Manpower planning06:57
199607:29
The ancestors of Data Mining09:20
KDD Definition09:43
The KDD process09:56
1997: Adaptive System Management10:25
Analysis of IT infrastructures using knowledge discovery12:15
Monitoring in the system12:25
Refining thresholds13:32
Prediction of failure13:58
Patents14:01
Experiments14:39
Helpdesk calls application15:55
Decision tree Scarlos performance16:23
Paging space develops polynomially17:13
CPU Idle Trend17:34
1998-200518:03
RoboSail19:20
Pole Balancing20:05
Ship Balancing (1)20:26
Ship Balancing (2)21:35
Ship Balancing (3)22:16
Robotician - robosail systems BV (1)22:20
Robotician - robosail systems BV (2)22:39
Robotician - robosail systems BV (3)23:19
Mining for Sensor Calibration (1)24:44
Mining for Sensor Calibration (2)25:21
Weka Knowledge Explorer25:22
2003-2009 Adaptive Information Disclosure (AID) participating in the VL-e (Virtual lav for e-Science) project25:39
Adaptive Information Disclosure26:59
The AIDA toolbox28:02
2010-2015: D2S: From Data to Semantics28:30
How do we speed up the transfer of scientific data, information, knowledge from a research paper into actionable form?28:54
Smart content at Elsevier & Philips29:21
Looking Forward (1)29:53
Looking Forward (2)29:59
Looking Forward (3)32:32
Research cycle (1)32:55
Research cycle (2)33:48
Science in the 21st century34:18
Research objective facticity project34:33
Noise35:11
JPEG File size with noise added35:30
Between order and chaos: facticity35:38
Facticity (Questions)36:02
Turing two-part code compression37:24
Definition of facticity: the amount of selfdescriptive information in a dataset37:49
Crash Course Complexity Theory37:56
Existing complexity measures39:08
Atlas of Complexity: Comparative Analysis of Metrics & Datasets (1)39:37
Atlas of Complexity: Comparative Analysis of Metrics & Datasets (1)41:35
Research Questions43:05
Medically relevant model of an individual human being43:41
Research cycle (3)44:00
Research Methodology Proposal I (1)44:07
Research Methodology Proposal I (2)44:28
Powerlaws44:57
Problems!45:09
Research Methodology Proposal II (1)45:41
Research Methodology Proposal II (2)46:12
VC dimension46:19
Problems again!!46:44
12,5Mb!!47:19
Information: Some numbers47:21
Empirical Incompleteness theory48:22
Research cycle (4)48:45
Mutual model information The classical view48:46
Crash Course Complexity Theory49:32
Mutual model information? Computational View49:42
Any two unrelated human beings differ by about 3 million distinct DNA variants.49:51
Loading model ...50:42
Breaking a long tradition51:31
Conclusions52:26