AI Tool

AI Tool has functions for creating data frames by extracting data sequences from analysis results and image files obtained by OCTA's simulation engines, and for analyzing data from data frames using machine learning and deep learning. Not only simulation data, but also experimental data and images can be used as data.
The results obtained by using OCTA's simulation engine (UDF data) are stored in a complex array structure of multiple arrays and nested arrays, but by using the data frame creation function, they can be converted into two-dimensional array data (numerical data) that can be used in machine learning. It is also possible to process the calculation results to calculate the necessary properties and create a data frame. In addition to OCTA simulation result files, microscope image files and CSV files are also supported.
The data analysis function provides templates for machine learning and deep learning, so machine learning and deep learning can be performed through GUI operations without writing scripts. For machine learning, Scikit-learn can be used, and for deep learning, TensorFlow/Keras can be used.

Basic functions

  • - Loading data
    Not only simulation and experimental data, but also CSV data and image data can be loaded and data framed.
  • - Data frame creation from simulation results (UDF)
    The results themselves, such as stresses and volume fractions, can be data-framed.
    Script (Action) can be used to process and frame data.
  • - Data frame creation from electron microscope (SEM, TEM, etc.) and simulation result images.
  • - Editing the created data frame
    Duplicate or delete data, merge multiple data, save in CSV, etc.
  • - Machine learning functions
    Interfaces for machine learning with Scikit-learn
    Supports multiple models such as Linear Regression, SVR, Logistic Regression, Random Forest, etc., and various execution parameter settings.
  • - Deep learning functions
    Interface for deep learning with Tensorflow/Keras
    Support for assembling layers of models and setting various execution parameters.
  • - Execution of python scripts
    Supports scripted data processing, data frame integration, and manipulation of machine learning and deep learning models.