Systems and methods for providing a self-learning estimation advisor

A method, comprising:

• receiving, by a processor in a system, from a user interacting with the system, a first document and a second document, wherein an actual value is associated with the second document;
• determining, by the processor, a set of variables for each of the first document and the second document, wherein the determining is based on a statistical analysis of historical data;
• segregating the set of variables into a first set of variables corresponding to a categorical variables and a second set of variables corresponding to numerical variables, wherein the categorical variables are converted into binary codes using an encoding technique, and wherein the numerical variables are normalized using a normalization technique;
• training, using a transformation module, an artificial neural network (ANN) model by performing linear transformation on the numerical variables, wherein the ANN model references a deep learning method including algorithms based on brain function, wherein the ANN model provides a relationship between inputs and outputs to discover a new pattern of the set of variables;
• estimating, by the processor, using a trained ANN model, a cost associated with each of the first document and the second document based on the segregated set of variables, wherein the cost associated with the first document comprises an estimated freight cost for the first document, and wherein the cost associated with the second document comprises an estimated true value for the second document;
• and
• providing the estimated cost associated with the first document and the second document associated with the second document to the user, via a user interface, interacting with the system.

A non-transitory computer-readable medium comprising machine-readable instructions that are executable by a processor to:

• receive, from a user interacting with the system, a first document and a second document;
• determine a set of variables for each of the first document and the second document, wherein the set of variables are determined based on a statistical analysis of historical data;
• segregate the set of variables into a first set of variables corresponding to a categorical variables and a second set of variables corresponding to numerical variables, wherein the categorical variables are converted into binary codes using an encoding technique, and wherein the numerical variables are normalized using a normalization technique;
• training, using a transformation module, an artificial neural network (ANN) model by performing linear transformation on the numerical variables, wherein the ANN model references a deep learning method including algorithms based on brain function, wherein the ANN model provides a relationship between inputs and outputs to discover a new pattern of the set of variables;
• estimate, using a trained ANN model, a cost associated with each of the first document and the second document based on the segregated set of variables, wherein the cost associated with the first document comprises an estimated freight cost for the first document, and wherein the cost associated with the second document comprises an estimated true value for the second document;
• and
• provide the estimated cost associated with the first document and the second document associated with the second document to the user, via a user interface, interacting with the system.