[Example] Support for interpreting machine learning models using XAI.

Implemented PFI, PD, ICE, and SHAP! Output feature importance and prediction reasons, and interpreted them!

In recent years, while highly accurate machine learning models have been developed, there is a problem of increasing black-box nature of these models. Since it is difficult to use them for decision-making in their black-box state, we aim to demonstrate the rationale behind predictions by using XAI. Our company has implemented model construction and XAI (PFI, PD, ICE, SHAP). We compiled the analysis results and the interpretations provided by XAI into a report and submitted it. 【Case Overview】 ■ Issues - In recent years, while highly accurate machine learning models have been developed, there is a problem of increasing black-box nature of these models. - It is difficult to use them for decision-making in their black-box state, so we want to demonstrate the rationale behind predictions by using XAI. ■ Work Content - Model construction and implementation of XAI (PFI, PD, ICE, SHAP) - Interpretation of model prediction results using XAI *For more details, please refer to the PDF document or feel free to contact us.

Embedded system design serviceOther contract services