These files are based on Tensorflow website tutorials: https://www.tensorflow.org/tutorials/estimator/premade, https://www.tensorflow.org/tutorials/images/cnn,
LogisticRegression_TF.py - using LinearClassifier I got 0.76 accuracy score on 'titanic' dataset
Classification_TF.py - using DNNClassifier the accuracy score is 0.80 on 'iris' dataset
NeuralNetwork_TF.py - Making predictions with feedforward NN from keras API, on MNIST Fashion dataset: loss: 0.34896257519721985 accuracy: 0.8816999793052673
Convolutional_NN_TF.py - Making predictions with convolutional+dense arch. NN's from keras API, on CIFAR dataset: loss: 0.871101438999176 accuracy: 0.7214000225067139
Pretrained_CNN_TF.py - Binary prediction on dogs/cats dataset using transfer learning of MobileNet V2 model and combine it in a custom model. After training, the model has: loss: 0.051920127123594284 accuracy: 0.9815133213996887
Recurrent_CNN_TF.py - Binary prediction on imdb reviews dataset using RNN specific layers for sentiment analysis. After training a simple RNN, the metrics are: loss: 0.2984626293182373 accuracy: 0.8804399967193604