Systems and methods to generate data messages indicating a probability of execution for data transaction objects using machine learning
A computer system includes a transceiver that receives over a data communications network different types of input data and multiple data transaction objects from multiple source nodes. A pre-processor processes the different types of input data and the data transaction objects to generate an input data structure. Based on the input data structure, one or more predictive machine learning models is trained and used to predict a probability of execution of each of the data transaction objects at a future execution time. Output data messages are then generated for transmission by the transceiver over the data communications network indicating the probability of execution for at least one of the data transaction objects at the future execution time.
1 . A distributed computer system to optimize, using predictive machine learning, allocation of computer and data communication resources for data transaction objects that are likely to be executed at a future execution time, comprising:
a transceiver configured to receive over a data communications network different types of input data and data transaction objects from multiple source nodes communicating with the data communications network;
a processing system that includes at least one hardware processor, the processing system configured to:
process the different types of input data and the data transaction objects to generate an input data structure for each of the data transaction objects;
based on the input data structure, predict, using one or more predictive machine learning models, a probability of execution of each of the data transaction objects at a future execution time;
allocating computer and data communication resources for data transaction objects that have a predicted probability of execution at the future execution time that equals or exceeds a predetermined probability threshold and avoiding allocating computer and data communication resources for data transaction objects that have a predicted probability of execution at the future execution time that is less than the predetermined probability threshold;
generate output data messages for transmission by the transceiver over the data communications network indicating the probability of execution for at least one of the data transaction objects at the future execution time;
monitoring and identifying changes that affect the data transaction objects; and
adapting the allocation of computer and data communication resources for data transaction objects based on the changes to reduce an amount of data communicated over data communication networks, lower consumption of other computer system resources, and improve performance of the computer system.
2 . The computer system in claim 1 , wherein each of the data transaction objects includes one or more conditions, and wherein the probability of execution for the at least one of the data transaction objects includes a probability of meetingsatisfying the one or more conditions associated with the one of the data transaction objects.
3 . The computer system in claim 1 , wherein the input data structure includes a combination of two or more of the different types of input data.
4 . The computer system in claim 1 , wherein the different types of input data have different data formats, and wherein the input data structure includes a vector format, a tabular format, and/or a hierarchical format.
5 . The computer system in claim 1 , wherein the generating includes generating an output data message indicating a probability of execution for one of the data transaction objects at the future execution time as a real time response to receiving the one data transaction object from a source node.
6 . The computer system in claim 1 , wherein the output data messages are generated in batches with each batch indicating a probability of execution for multiple ones of the data transaction objects at the future execution time, and wherein the batches are generated periodically.
7 . The computer system in claim 1 , wherein the output data messages are in tabular format, text format, and/or a hierarchical format.
8 . The computer system in claim 1 , wherein the output data messages include fields that indicate multiple parameters and multiple conditions for each of the data transaction objects.
9 . The computer system in claim 1 , wherein the one or more predictive machine learning models includes a gradient boost prediction model.
10 . The computer system in claim 1 , wherein the processing system is configured to training the one or more predictive machine learning models by the following steps:
(i) adding a base predictive model with a further predictive model to generate a current base predictive model, wherein the further predictive model is based on errors of the current base predictive model predicting execution of the data transaction objects at the future execution time as compared to actual execution of the data transaction objects at the future execution time, and
(ii) repeating step (i).
11 . The computer system in claim 10 , wherein the processing system is configured to repeat step (i) until the errors are below a predetermined error threshold, the number of iterations reach a predetermined number, or a decrease in the errors for a current repetition as compared to the errors for one or more prior repetitions is less than a threshold.
12 . The computer system in claim 10 , wherein each of the predictive machine learning models includes a decision tree or a logistic regression.
13 . A method, comprising:
receiving by a transceiver over a data communications network different types of input data and data transaction objects received from multiple source nodes communicating with the data communications network;
executing, by a processing system that includes at least one hardware processor, instructions stored in memory and, in accordance with the instructions, performing operations that include:
(a) processing the different types of input data and the data transaction objects to generate an input data structure for each of the data transaction objects;
(b) based on the input data structure, predicting, using one or more predictive machine learning models, a probability of execution of each of the data transaction objects at a future execution time;
(c) using the predicted probability of execution of each of the data transaction objects to optimize allocation of computer and data communication resources including allocating computer and data communication resources for data transaction objects that have a predicted probability of execution at the future execution time that equals or exceeds a predetermined probability threshold and avoiding allocating computer and data communication resources for data transaction objects that have a predicted probability of execution at the future execution time that is less than the predetermined probability threshold;
(d) generating output data messages for transmission by the transceiver over the data communications network indicating the probability of execution for at least one of the data transaction objects at the future execution time;
(e) monitoring and identifying changes that affect the data transaction objects; and
(f) adapting the allocation of computer and data communication resources for data transaction objects based on the changes to reduce an amount of data communicated over data communication networks, lower consumption of other computer system resources, and improve performance of the computer system.
14 . The method in claim 13 , wherein each of the data transaction objects includes one or more conditions, and wherein the probability of execution for at least one of the data transaction objects includes a probability of meeting the one or more conditions associated with the one of the data transaction objects.
15 . The method in claim 13 , wherein the different types of input data have different data formats, and wherein the input data structure includes a vector format, a tabular format, and/or a hierarchical format.
16 . The method in claim 13 , wherein the generating includes generating, by the processing system, an output data message indicating a probability of execution for one of the data transaction objects at the future execution time as a real time response to receiving the one data transaction object from a source node.
17 . The method in claim 13 , wherein the output data messages are generated, by the processing system, in batches with each batch indicating probabilities of execution for multiple ones of the data transaction objects at the future execution time, and wherein the batches are generated periodically.
18 . The method in claim 13 , wherein the output data messages include fields that indicate multiple parameters and multiple conditions for each of the data transaction objects.
19 . The method in claim 13 , further comprising training the one or more predictive machine learning models by the processing system performing the following steps:
(g) adding a base predictive model with a further predictive model to generate a current base predictive model, wherein the further predictive model is based on errors of the current base predictive model predicting execution of the data transaction objects at the future execution time as compared to actual execution of the data transaction objects at the future execution time, and
(h) repeating step (g).
20 . The method in claim 19 , wherein step (g) is repeated, by the processing system, until the errors are below a predetermined error threshold, the number of iterations reach a predetermined number, or a decrease in the errors for a current repetition as compared to the errors for one or more prior repetitions is less than a threshold.
21 . A non-transitory, computer-readable medium encoded with instructions that, when executed by at least one hardware processor in a distributed computer system, cause the at least one hardware processor to receive over a data communications network different types of input data and data transaction objects from multiple source nodes communicating with the data communications network and to perform operations comprising:
(a) processing the different types of input data and the data transaction objects to generate an input data structure for each of the data transaction objects;
(b) based on the input data structure, predicting, using one or more predictive machine learning models, a probability of execution of each of the data transaction objects at a future execution time;
(c) using the predicted probability of execution of each of the data transaction objects to optimize allocation of computer and data communication resources including allocating computer and data communication resources for data transaction objects that have a predicted probability of execution at the future execution time that equals or exceeds a predetermined probability threshold and avoiding allocating computer and data communication resources for data transaction objects that have a predicted probability of execution at the future execution time that is less than the predetermined probability threshold;
(d) generating output data messages for transmission by the transceiver over the data communications network indicating the probability of execution for at least one of the data transaction objects at the future execution time;
(e) monitoring and identifying changes that affect the data transaction objects; and
(f) adapting the allocation of computer and data communication resources for data transaction objects based on the changes to reduce an amount of data communicated over data communication networks, lower consumption of other computer system resources, and improve performance of the computer system.