Stochastic Threshold Model Trees: A Tree-Based Ensemble Method for Dealing with Extrapolation
This work addresses the challenge of extrapolation in chemistry for new material development, but it is incremental as it builds on existing tree-based methods with limited overall accuracy gains.
The paper tackles the problem of predicting properties of unknown compounds in extrapolation areas where no known data exists, proposing Stochastic Threshold Model Trees (STMT) to maintain interpolation accuracy while reflecting data trends, with experiments showing notable prediction improvement for one compound in real data.
In the field of chemistry, there have been many attempts to predict the properties of unknown compounds from statistical models constructed using machine learning. In an area where many known compounds are present (the interpolation area), an accurate model can be constructed. In contrast, data in areas where there are no known compounds (the extrapolation area) are generally difficult to predict. However, in the development of new materials, it is desirable to search this extrapolation area and discover compounds with unprecedented physical properties. In this paper, we propose Stochastic Threshold Model Trees (STMT), an extrapolation method that reflects the trend of the data, while maintaining the accuracy of conventional interpolation methods. The behavior of STMT is confirmed through experiments using both artificial and real data. In the case of the real data, although there is no significant overall improvement in accuracy, there is one compound for which the prediction accuracy is notably improved, suggesting that STMT reflects the data trends in the extrapolation area. We believe that the proposed method will contribute to more efficient searches in situations such as new material development.