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

Machine learning for environmental and life sciences

Published on 2019-03-27250 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

Presentation

Machine Learning for Environmental & Life Sciences00:00
What is artificial intelligence?10:27:06
Machine Learning29:13:46
Data Science47:15:35
The most popular data science topic81:35:50
Machine Learning of Interpretable Models109:14:44
The basic Machine Learning task: Predictive modeling114:59:39
An example task of Predictive Modelling: Medical diagnosis122:28:45
Example task: Descriptive vars.; Biomarkers for Alzheimer’s130:15:59
Example: Decision tree for diagnosis137:26:26
Another example of single-target predictive modeling (classification) - 1141:15:05
Another example of single-target predictive modeling (classification) - 2153:07:35
What is a decision tree?161:00:34
Making a Prediction with a Decision Tree173:01:38
Machine Learning of Decision Trees195:58:06
Top-Down Induction of Decision Trees250:38:56
TDIDT Illustrated251:43:07
Mining Big and Complex Data:Dimensions of Complexity283:38:12
Mining Big and Complex Data295:52:24
Big Data: Variety - Structured Input - 1318:42:17
Big Data: Variety - Structured Input - 2329:43:01
Predictive modeling: Structured output338:00:09
Big Data: Volume & Velocity343:48:07
Data streams: Regression353:11:29
Semi-supervised learning: Classification and regression358:56:51
Data in context: Spatio-temporal, network383:10:36
The Different Tasks of Multi-Target Prediction398:23:22
Weather prediction403:17:20
Multi-target prediction417:57:06
Example MTR task: Target vars.; Clinical scores for Alzheimer’s425:27:43
Example MTR model439:58:38
Multi-Target Classification & Multi-Label Classification450:51:52
Multi-Label Classification Example461:22:22
Hierarchical multi-label classification473:21:37
Hierarchical multi-label classif.476:52:38
Hierarchical multi-label classification: Another example500:02:14
The hierarchy can be a tree or a DAG506:00:55
Hierarchical structure on target space for the ADNI dataset514:22:15
Mining Big and Complex Data:Combining Complexities527:03:07
Data streams: (MT) Regression535:34:42
Network +SOP: HMC537:53:44
Even more complicated tasks560:46:02
Predictive Clustering for Multi-Target Prediction564:44:52
Clustering574:43:32
Example predictive clustering tree588:09:00
Top-down induction of PCTs613:11:42
Predictive clustering647:43:09
Selecting the best test in a PCT661:41:11
Multi-target regression664:36:42
Multi-target classification667:17:38
Ensembles of PCTs671:42:57
SSL+SOP: Incomplete Annotations689:01:57
Learning PCTs693:35:08
Relating the Environment and the Biota697:22:13
Environment <-> Biota713:01:33
Habitat modeling717:48:31
Predicting species composition720:08:05
Predicting community structure722:08:09
Slovenian rivers724:34:58
Danish farms: Soil Microarthropods732:54:17
Victoria, AustraliaVegetation734:19:50
Slovenian rivers: Species comp.741:48:18
Slovenian rivers: Habitat models749:07:05
Slovenian rivers: Community struc.752:49:39
Community structure: Overall results757:17:02
Predicting Gene Functions814:09:43
Predicting gene functions817:05:15
Predicting Gene Functions in Bacterial Genomes (RBI+JSI)826:20:04
GFP Pipeline843:47:48
Different Features Sets for GFP857:46:22
Gene Function Prediction: Predictive Performance867:30:29
Metagenome Phyletic Profiles880:40:51
Multi-Target Prediction for Virtual Compound Screening902:30:56
Virtual compound screening906:58:35
Host-targeted Drugs for MTB (Tuberculosis) and STM (Salmonella)939:00:51
MTB&STM: Host-targeted Drugs951:55:50
MTB&STM: Host-targeted Drugs The Data Analysis Workflow961:58:38
MTB&STM: Host-targeted DrugsResults977:53:43
Analyzing data from High-contents Screens983:00:11
HTS: Modulating fibroblast to myofibroblast transition992:41:15
Hits in the HTS screen995:29:07
Reducing fibrosis in myocardial infarction998:11:05
Testing the predictions1001:58:23
Spring school in Bled in May1014:15:49
Conclusions1027:11:04