Uncertainty-Aware prediction model
| Main Author: | Mohammad Heydari |
|---|---|
| Format: | info Lainnya eJournal |
| Bahasa: | eng |
| Terbitan: |
, 2020
|
| Subjects: | |
| Online Access: |
https://zenodo.org/record/3755598 |
Daftar Isi:
- 1- Uncertainty.ipynp: Python source code file in which prediction models are trained and built using a dataset (LD.xlsx). The code uses pandas library to import the dataset and build prediction models (Decision Tree, SVM, Logistic Regression, Naive Bayes, AdaBoost, Random Forest, K-NN, MLP, Gradient Boost and Voting Classifier) using sklearn library. All models are trained/built using 10-fold cross validation method. Moreover, a Gaussian noise function is used to add noise to the dataset. The code saves all prediction models and measures the performance of the models using accuracy, precission, recall and F1 metrics. It also draws ROC curves for each prediction model. 2- AdaBoostClassifier.sav: Designated prediction model (AdaBoost) 3- LD.xlsx: Low mobility dataset (Labeled) in Excel format consisting of uncertainty values for three attributes: Time, Location and Credential.