A complete end-to-end machine learning portal that covers processes starting from model training to model predicting results using FastAPI.
- Dataset: Set default dataset from the list that will be used for training and building the models.
- Train: Start the training process
- Prediction: Upload test dataset to get predictions.
Save sample test data (sample_test.csv) for testing the prediction module.
- Data Preprocessing: Logs generated during preprocessing process.
- Training: Logs generated during training process.
- Prediction: Logs generated during prediction process.
- Import/upload dataset
- Preprocessing
- Categorical features cleaning
- Handling missing value (categorical)
- Handling missing value (numeric)
- Encoding cateogrical feature
- Over sampling
- Clustering
- Train-Test split
- Model selection
- Hyper parameter tuning
- Saving best model
- Download prediction results
