Classification of AML
Definitions
Pearson: The Pearson correlation between the ideal confidence values (that is, 1 for the twenty AML patients and 0 for other subjects) and the confidence of the predictions for each patient. This measure determined an order amongst the top performing teams whose precision and recall were 1.
Prec: The Precision of the predictions, defined as the fraction of correct AML patients amongst the first 20 predictions.
Rec: The Recall of the predictions, defined as the proportion of AML patients in the first 20 predictions out of all the AML patients in the cohort.
MCC: The Matthews Correlation Coefficient is a measure of the quality of binary classifications. It takes into account true and false positives and negatives and is generally regarded as a balanced measure which can be used even if the classes are of very different sizes. For its mathematical definition see http://en.wikipedia.org/wiki/Matthews_correlation_coefficient
JSC: The Jaccard Similarity Coefficient measures similarity between two sample sets, and is defined as the size of the intersection divided by the size of the union of the sample sets. In this case, set 1 is composed by the first 20 subjects predicted to be affected to AML, and set 2 is the set of all the twenty AML patients. (See http://en.wikipedia.org/wiki/Jaccard_index)
Score: The average of the Prec, Rec, MCC and Jaccard coefficient. The maximum possible value is 1 for a perfect classification.
Rank: The rank of the teams according to score.
Rank Among Best Performers: The ranking among the top performing teams (Rank=1), based on their Pearson metric. Some teams were concerned that this is a somehow arbitrary measure of performance. To read about the divergent opinions on this issue please follow the thread in the Dicussion Forum at http://www.the-dream-project.org/forum/challenge-4-pearson-correlation
If you participated in this challenge, login to de-anonymize your team
|
Team |
Pearson |
Prec |
Rec |
MCC |
JSC |
Score |
Rank |
Rank Among Best Performers |
|
team21 |
0.9952 |
1.00 |
1.00 |
1.00 |
1.00 |
1.00 |
1 |
1 |
|
BCB |
0.9891 |
1.00 |
1.00 |
1.00 |
1.00 |
1.00 |
1 |
2 |
|
Team #247 |
0.9703 |
1.00 |
1.00 |
1.00 |
1.00 |
1.00 |
1 |
3 |
|
Team #56 |
0.9663 |
1.00 |
1.00 |
1.00 |
1.00 |
1.00 |
1 |
4 |
|
Team #251 |
0.9656 |
1.00 |
1.00 |
1.00 |
1.00 |
1.00 |
1 |
5 |
|
Team #167 |
0.9617 |
1.00 |
1.00 |
1.00 |
1.00 |
1.00 |
1 |
6 |
|
Team #144 |
0.8804 |
1.00 |
1.00 |
1.00 |
1.00 |
1.00 |
1 |
7 |
|
Team #252 |
0.8727 |
1.00 |
1.00 |
1.00 |
1.00 |
1.00 |
1 |
8 |
|
Team #181 |
0.9666 |
0.95 |
0.95 |
0.94 |
0.90 |
0.94 |
9 |
- |
|
Team #262 |
0.9539 |
0.95 |
0.95 |
0.94 |
0.90 |
0.94 |
9 |
- |
|
Team #208 |
0.9438 |
0.95 |
0.95 |
0.94 |
0.90 |
0.94 |
9 |
- |
|
Team #61 |
0.8927 |
0.95 |
0.95 |
0.94 |
0.90 |
0.94 |
9 |
- |
|
Team #21 |
0.6455 |
0.95 |
0.95 |
0.94 |
0.90 |
0.94 |
9 |
- |
|
Team #36 |
0.9269 |
0.90 |
0.90 |
0.89 |
0.82 |
0.88 |
14 |
- |
|
Team #107 |
0.8869 |
0.90 |
0.90 |
0.89 |
0.82 |
0.88 |
14 |
- |
|
Team #176 |
0.8695 |
0.85 |
0.85 |
0.83 |
0.74 |
0.82 |
16 |
- |
|
Team #261 |
0.6802 |
0.80 |
0.80 |
0.78 |
0.67 |
0.76 |
17 |
- |