Model Creating Pipeline
The model training process in Neuton contains the following steps:
Step 1: Virtual Machine (VM) creation
A Virtual Machine (VM) is automatically provisioned to perform the following tasks: data preprocessing, feature engineering, model training, validation, and preparation of the C Library for predictions. This process is fully automated and does not require any user input or action.
Step 2: Data Preparation
Digital Signal Processing (DSP) for Sensor Data
This option makes data suitable for creating models. DSP consists of windowing, feature extraction, and feature selection. The user can flexibly choose the needed settings.
Data Preprocessing and Feature Engineering for Tabular Data
Data preprocessing and feature engineering for Tabular Data are fully automated. The platform itself determines the necessary data transformations.
Conversion of WAV Files for Audio Data
Converting audio files to a CSV format includes building a spectrogram and performing other transformations of audio signals according to the parameters specified by the user.
Step 3: Model training and validation
During this step, our proprietary neural network algorithm, Neuton, will automatically create a neural network architecture to achieve the best possible score on the validation data (measured by the metric selected by the user).
Once model training is complete, a downloadable archive is generated. It contains everything you need to embed the model into an MCU or perform a prediction in the user's infrastructure. It is also possible to perform predictions directly on the platform.
Step 4: Virtual Machine termination
Upon completion of training, the virtual machine will be automatically deprovisioned.