1. Quick start

Before you get started with FEDOT.Industrial, it is must be installed (refer to the installation instructions).

Using the API

FEDOT.Industrial provides a simple API for various tasks, including classification, regression, time series forecasting, and anomaly detection.

Steps:

1. Import the FedotIndustrial class:

from fedot_ind.api.main import FedotIndustrial

2. Initialize the FedotIndustrial object: Specify the type of modeling task you want to perform. The object provides methods for fitting a model, making predictions, and evaluating performance.

• fit(): This method performs feature extraction, optimization, and returns the resulting model pipeline.
• predict(): Use this method to predict target values for new data using the fitted model.
• get_metrics(): This method estimates the quality of predictions using chosen metrics.

Both NumPy arrays and Pandas DataFrames can be used for input data. In the following example, x_train, y_train, and x_test are NumPy arrays:

model = FedotIndustrial(problem='classification',
                        metric='f1',
                        timeout=5)

model.fit((x_train, y_train))
prediction = model.predict((x_test, y_test))
metrics = model.get_metrics(target=y_test,
                            metric_names=['f1', 'accuracy'],
                            rounding_order=3)

DataLoader tool

For time series classification tasks, you can utilize the DataLoader class to download data from the UCR/UEA archive:

from fedot_ind.tools.loader import DataLoader

loader = DataLoader(dataset_name='ECG200')
train_data, test_data = loader.load_data()

Using Your Own Data

If you have your own data in .ts, .tsv, or .arff format, specify the folder path when initializing the DataLoader:

loader = DataLoader(dataset_name='YourDatasetName', folder_path='path/to/folder')
train_data, test_data = loader.load_data()