activation function

Methods and means for real-time object recognition accuracy increase in video images on iOS mobile platform

As a result of the analytical review, it was established that the family of Yolo models is a promising area of search and recognition of objects. However, existing implementations do not support the ability to run the model on the iOS platform. To achieve these goals, a comprehensive scalable conversion system has been developed to improve the recognition accuracy of arbitrary models based on the Docker system. The method of improvement is to add a layer with the Mish activation function to the original model. The method of conversion is to quickly convert any Yolo model to CoreML format.

Прогнозування відмов програмного забезпечення з використанням нейронної мережі на основі радіально-базисних функцій

In this paper the radial-basis neural network was used for software failures prediction. The influence of activation function of the RBF neural net on the learning efficiency and software failures prediction is studied. It is shown that the optimal activation function is Inverse Multiquadric with 10 neurons in the input layer and 30 neurons in the hidden one (square of Pearson correlation coefficient is 0.997 and mean deviation is 14.4).