Prediction model for determining whether feature vector of data in each of multiple input sequences should be added to that of the other data in the sequence

A computer program product for predicting properties of an object from sequence data as a set of feature vectors describing the object, the computer program product comprising a non-transitory computer readable storage medium having program instructions embodied therewith, the program instructions executable by a computer to cause the computer to perform a method comprising:

• generating, by a hardware processor, a probabilistic prediction model y* predicting an objective variable y in a sequence-to-sequence feature vector copying process determining whether a respective feature vector of the sequence data in multiple input data sequences is added to sequence data in other input data sequences using sparse Bayesian criterion with Automatic Relevance Determination (ARD) and variational approximation;
• configuring, by the hardware processor, the probabilistic prediction model y* to (i) assign a prediction function of multiple prediction functions for each of the feature vectors extracted from the sequence data, (ii) identify a relationship between a t-th feature vector in an i-th data and the objective variable y, and (iii) identify similarities of relationships between the feature vectors and the objective variable y by learning parameters for the similarities that include a strength of co-occurrence of the prediction function in the t-th vector and a t-1-th vector using the sequence data;
• identifying, by the hardware processor using the probabilistic prediction model y*, a sequence length which is variable between the multiple input data sequences;
• predicting, by the hardware processor, the objective variable y as a chemical property of the object based on the probabilistic prediction model y*;
• and
• forming, by a manufacturing system coupled to the hardware processor, a new compound based on the prediction of the objective variable y as a constituent element of the new compound.

A computer processing system for predicting properties of an object from sequence data as a set of feature vectors describing the object, the computer processing system comprising:

• a memory for storing program code;
• and
• a hardware processor for executing the program code to:
• generate a probabilistic prediction model y* predicting an objective variable y in a sequence-to-sequence feature vector copying process determining whether a respective feature vector of the sequence data in multiple input data sequences is added to sequence data in other input data sequences using sparse Bayesian criterion with Automatic Relevance Determination (ARD) and variational approximation;
• configure the probabilistic prediction model y* to (i) assign a prediction function of multiple prediction functions for each of the feature vectors extracted from the sequence data, (ii) identify a relationship between a t-th feature vector in an i-th data and the objective variable y, and (iii) identify similarities of relationships between the feature vectors and the objective variable y by learning parameters for the similarities that include a strength of co-occurrence of the prediction function in the t-th vector and a t-1-th vector using the sequence data;
• identify, using the probabilistic prediction model y*, a sequence length which is variable between the multiple input data sequences;
• predict the objective variable y as a chemical property of the object based on the probabilistic prediction model y*;
• and
• a manufacturing system coupled to the hardware processor for forming a new compound based on the prediction of the objective variable y as a constituent element of the new compound.