Groupe1 Groupe2 Accuracy(F-Measure) How well can it deal with missing values? How much does the prediction performance conserve with missing values? How well can it deal with correlated attributes? Deal with numerical data? How well can it deal with errors or noises? How much is it overfitted? How scalable is it? Is it incremental? How explainable is it? How transparent is it? Pertinent Hyper-parameters and values How deterministic is it? Quantity of data that can deal with? Number of features that can deal with? How much knowledge about features does it accept? How well can it deal with linearly separated data? How much time can it take during training? How much time can it take during prediction? How much memory does it need during training? How much memory does it need during prediction? Type of training How much imbalance can this algorithm treat? How well can it deal with normally distributed data? Add more info about the algorithm How much is it distributable? How much can be used in federated learning? Propose data pre-processing pipelines for the ML algorithm Deal with categorical data? Can it deal with Bagging (bootstrap aggregation) ? Can it deal with Boosting (adaboost)? Can it deal with Gradient boosting ? Is this algorithm use stochastic gradient descent ?
KNN ShallowModelSuperviseLearning
Naïve Bayes ShallowModelSuperviseLearning
Bayesian Network ShallowModelSuperviseLearning
Random Forest ShallowModelSuperviseLearning
Neural network ShallowModelSuperviseLearning
SVM ShallowModelSuperviseLearning
Logistic regression ShallowModelSuperviseLearning
CNN (convolutional neural network) DeepLearningModelSuperviseLearning
Gaussien process classifier (GPR) ShallowModelSuperviseLearning
Linear discriminant analysis (LDA) ShallowModelSuperviseLearning
Decision Trees ShallowModelSuperviseLearning
Transformer neural networks DeepLearningModelSuperviseLearning
Long short-term memory (LSTM) DeepLearningModelSuperviseLearning
K-means ShallowModelUnsepervisedLearning
Gated Recurrent Units DeepLearningModelSuperviseLearning
Bidirectional RNNs DeepLearningModelSuperviseLearning
Generative Adversarial Network DeepLearningModelUnsepervisedLearning
Deep Belief Networks DeepLearningModelUnsepervisedLearning
Autoencoder DeepLearningModelUnsepervisedLearning
Model-based RL DeepLearningModelReinforcementLearning
Q-Learning DeepLearningModelReinforcementLearning
Temporal differences DeepLearningModelReinforcementLearning