There are 0 repository under false-negative topic.
RNAchallenge dataset for the classification of protein-coding and non-coding RNAs
[IJHCS] UTA7: a dataset of rates (BI-RADS) provided by clinicians resulted from classifying the given medical images for breast cancer diagnosis.
With a precision of 86% and model's CAP curve showing an accuracy of 100%! This means it is capable of correctly predicting 100% of patients with a heart disease after processing 50% of the data. The model's performance is "Too Good to be True"! However, with Train accuracy = 86% and Test accuracy = 82%, there is no visible sign of overfitting.
A tool reading ground truth data and detected object data provides visualization way to estimate position accuracy.
Machine learning for credit card default. Precision-recalls are calculated due to imbalanced data. Confusion matrices and test statistics are compared with each other based on Logit over and under-sampling methods, decision tree, SVM, ensemble learning using Random Forest, Ada Boost and Gradient Boosting. Easy Ensemble AdaBoost classifier appears to be the model of best fit for the given data.
Slides of a talk on deep learning, false negatives/positives and predator-prey interactions with R.