video thumbnail
Pause
Mute
Subtitles not available
Playback speed
0.25
0.5
0.75
1
1.25
1.5
1.75
2
Full screen

Multi-Target Prediction with Trees and Tree Ensembles

Published on 2019-06-28169 Views

Increasingly often, we need to learn predictive models from big or complex data, which may comprise many examples and many input/output dimensions. When more than one target variable has to be predict

Related categories

Presentation

Multi-target Prediction with Trees & Tree Ensembles00:00
The basic Machine Learning task: Predictive modeling01:09
An example task of Predictive Modelling: Medical diagnosis01:36
Example task: Descriptive vars.; Biomarkers for Alzheimer’s02:06
Example: Decision tree for diagnosis02:37
Another example of single-target predictive modeling (classification)03:02
Another example of single-target predictive modeling (classification) - 204:13
What is a decision tree?05:31
Making a Prediction with a Decision Tree05:54
Top-Down Induction of Decision Trees06:43
TDIDT Illustrated08:57
Mining Big and Complex Data: Dimensions of Complexity11:29
Big Data: Volume & Velocity13:56
Data streams: Regression15:12
Big Data: Variety - Structured Input15:28
Big Data: Variety - Structured Input - 216:34
Semi-supervised learning: Classification and regression18:10
Data in context: Spatio-temporal, network19:27
The Different Tasks of Multi-Target Prediction20:39
Weather prediction20:52
Multi-target prediction21:58
Example MTR task: Target vars.; Clinical scores for Alzheimer’s22:21
Example MTR model23:30
Multi-Target Classification & Multi-Label Classification24:13
Multi-Label Classification Example24:52
Hierarchical multi-label classification25:18
Hierarchical multi-label classification - 225:56
Hierarchical multi-target regression26:44
Mining Big and Complex Data: Combining Complexities28:02
SSL+SOP: Incomplete Annotations28:05
Data streams: (MT) Regression29:06
Network +SOP: HMC29:21
Predictive Clustering for Multi-Target Prediction29:55
Clustering30:16
Example predictive clustering tree31:12
Top-down induction of PCTs32:48
Learning PCTs34:22
Learning PCTs - 235:05
Selecting the best test in a PCT36:52
Multi-target regression37:06
Ensembles of PCTs37:31
Relating the Environment and the Biota: From Habitat models to Community composition38:20
Environment <-> Biota38:42
Habitat modeling38:53
Predicting species composition39:18
Predicting community structure39:30
Community structure: Overall results40:09
SSL in MTP: Accuracy & interpretability42:19
Multi-Target Prediction for Virtual Compound Screening44:43
Virtual compound screening44:48
Host-targeted Drugs for MTB (Tuberculosis) and STM (Salmonella)45:33
MTB&STM: Host-targeted Drugs46:14
MTB&STM: Host-targeted Drugs The Data Analysis Workflow47:14
MTB&STM: Host-targeted Drugs Results47:58
Analyzing data from high-contents screens48:47
Reducing fibrosis in myocardial infarction49:07
Testing the predictions49:19
Conclusions49:38