azureml-training-tabular

Create a small data science application using Python and the 'azureml-training-tabular' package. This application will predict housing prices based on various features such as location, size, number of bedrooms, and more. Your goal is to build a fully functional mini-app that showcases the capabilities of 'azureml-training-tabular' for both automated machine learning (AutoML) and standalone model training.

### Project Steps:
1. **Data Collection**: Gather a dataset containing information about houses and their prices. You can use a publicly available dataset like the Boston Housing dataset or create your own dataset.
2. **Data Preprocessing**: Clean the dataset by handling missing values, converting categorical data into numerical format, and normalizing the data if necessary.
3. **Feature Engineering**: Utilize 'azureml-training-tabular' to perform feature engineering tasks such as adding new features or transforming existing ones to better capture the underlying patterns in the data.
4. **Model Training**: Use 'azureml-training-tabular' to train a model either through AutoML or manually select a model from the package and train it on your preprocessed dataset.
5. **Model Evaluation**: Evaluate the performance of your trained model using appropriate metrics such as Mean Absolute Error (MAE), Root Mean Squared Error (RMSE), or R-squared value.
6. **Scoring Code**: Implement scoring code to make predictions on new, unseen data.
7. **Deployment**: Although not mandatory, consider deploying your model using Azure Machine Learning services to showcase end-to-end functionality.

### Suggested Features:
- **Interactive Data Exploration Interface**: Allow users to explore different aspects of the dataset interactively before proceeding to model training.
- **Customizable Feature Engineering**: Provide options for users to customize feature engineering steps according to their needs.
- **Comparison of Models**: Automatically compare multiple models generated through AutoML to help users choose the best performing one.
- **User-Friendly Model Scoring**: Create a simple interface where users can input house characteristics and receive predicted prices.

### Utilization of 'azureml-training-tabular':
- Use the package's built-in ML models and featurizers for efficient and effective data processing and modeling.
- Leverage AutoML functionalities within the package to automate the model selection process and reduce manual effort.
- Integrate the provided scoring code to streamline the deployment and prediction phases of your application.

Your application should be designed to be user-friendly, interactive, and educational, demonstrating the power of 'azureml-training-tabular' in real-world scenarios.