Systems and methods for features engineering

1. A computer-implemented method for training one or more machine learning models for forecasting the demand for products at different locations, wherein the method uses features engineering and the method comprises the steps of:

receiving, by an ingestion module, internal signal data comprising columns and values;

receiving, by the ingestion module, external signal data comprising columns and values;

fusing data from the internal signal data and the external signal data, the fusing based on meta-data of each of the internal signal data and each of the external signal data;

generating a plurality of features based on all of the combinations of columns that match transformation inputs, the transformations forming part of a library of transformations;

applying the combinations of columns and transformations to the fused data to produce feature generated data that can be used to train machine learning models;

filtering the feature generated data down to a smaller set of columns that have enhanced predictive power;

selecting one or more features from the plurality of features, based on a predictive strength of each feature, to produce a set of selected features; and

training the one or more machine learning models using the set of selected features.

2. The computer-implemented method of claim 1 , wherein the external signal data is at least one of a weather signal and a financial signal.

3. The computer-implemented method of claim 1 , wherein:

at least one of the internal signal data and the external signal data includes a range; and

the at least one of the internal signal data and the external signal data is expanded to include one or more individual fields of the range.

4. A computing apparatus for training one or more machine learning models for forecasting the demand for products at different locations, wherein the computing apparatus uses features engineering, and the computing apparatus comprises:

a processor; and

a memory storing instructions that, when executed by the processor, configure the apparatus to:

receive, by an ingestion module, internal signal data comprising columns and values;

receive, by the ingestion module, external signal data comprising columns and values;

fuse data from the internal signal data and the external signal data, the fusing based on meta-data of each of the internal signal data and each of the external signal data;

generate a plurality of features based on all of the combinations of columns that match a transformation inputs, the transformations forming part of a library of transformations;

apply the combinations of columns and transformations to the fused data to produce feature generated data that can be used to train machine learning models;

filter the feature generated data down to a smaller set of columns that have enhanced predictive power;

select one or more features from the plurality of features, based on a predictive strength of each feature, to produce a set of selected features; and

train the one or more machine learning models by using the set of selected features.

5. The computing apparatus of claim 4 , wherein the external signal data is at least one of a weather signal and a financial signal.

6. The computing apparatus of claim 4 , wherein:

at least one of the internal signal data and the external signal data includes a range; and

the at least one of the internal signal data and the external signal data is expanded to include one or more individual fields of the range.

7. A non-transitory computer-readable storage medium for training one or more machine learning models for forecasting the demand for products at different locations, wherein the method uses features engineering and the computer-readable storage medium includes instructions that when executed by a computer, cause the computer to:

receive, by an ingestion module, internal signal data comprising columns and values;

receive, by the ingestion module, external signal data comprising columns and values;

fuse data from the internal signal data and the external signal data, the fusing based on meta-data of each of the internal signal data and each of the external signal data;

generate a plurality of features based on all of the combinations of columns that match transformation inputs, the transformations forming part of a library of transformations;

apply the combinations of columns and transformations to the fused data to produce feature generated data that can be used to train machine learning models;

filter the feature generated data down to a smaller set of columns that have enhanced predictive power;

select one or more features from the plurality of features, based on a predictive strength of each feature, to produce a set of selected features; and

train the one or more machine learning models by using the set of selected features.

8. The computer-readable storage medium of claim 7 , wherein the external signal data is at least one of a weather signal and a financial signal.

9. The computer-readable storage medium of claim 7 , wherein:

at least one of the internal signal data and the external signal data includes a range; and

the at least one of the internal signal data and the external signal data is expanded to include one or more individual fields of the range.

10. The computer-implemented method of claim 1 , wherein the method comprises:

training a plurality of machine learning models using the set of selected features;

evaluating the trained machine learning models to determine which of the trained machine learning models is the best trained machine learning model; and

executing the best trained machine learning model on processed data to provide a forecast of the demand for products at different locations.

11. The computing apparatus of claim 4 , wherein the instructions that, when executed by the processor, further configure the apparatus to:

train a plurality of machine learning models using the set of selected features;

evaluate the trained machine learning models to determine which of the trained machine learning models is the best trained machine learning model; and

execute the best trained machine learning model on processed data to provide a forecast of the demand for products at different locations.

12. The computer-readable storage medium of claim 7 , wherein the instructions further configure the computer to:

train a plurality of machine learning models using the set of selected features;

evaluate the trained machine learning models to determine which of the trained machine learning models is the best trained machine learning model; and

execute the best trained machine learning model on processed data to provide a forecast of the demand for products at different location.

13. A computer-implemented method for forecasting sales of a product at a location, the method comprising the steps of:

engineering, by the processor, one or more features; and

choosing, by the processor, a forecasting method, the forecast method based on a forecast request;

wherein:

engineering one or more features comprises:

receiving, by the processor, internal signal data comprising columns and values;

receiving, by the processor, external signal data comprising columns and values;

fusing, by the processor, data from the internal signal data and the external signal data, the fusing based on meta-data of each of the internal signal data and each of the external signal data;

generating, by the processor, a plurality of features based on all of the combinations of columns that match transformation inputs, the transformations forming part of a library of transformations;

applying, by the processor, the combinations of columns and transformations to the fused data to produce feature generated data that can be used to train machine learning models;

filtering, by the processor, the feature generated data down to a smaller set of columns that have enhanced predictive power; and

selecting, by the processor, one or more features from the plurality of features, based on a predictive strength of each feature, to produce a set of selected features; and

wherein choosing the forecasting method comprises:

training, by the processor, one or more machine learning models, using the set of selected features;

and either:

i) selecting a machine learning model from the one or more machine learning models to make the forecast; or

ii) retraining a previously-selected machine learning model from the one or more machine learning models and using the retrained model to make the forecast; or

iii) making the forecast based on the previously-selected machine learning model.

14. The computer-implemented method of claim 13 , wherein:

the forecast request is a first request;

the forecasting method selects the machine learning model to make the forecast, selecting comprising:

training the one or more machine learning models on a first portion of a data set;

validating the machine learning model on a second portion of the data set; and

retraining the machine learning model on a sum of the first portion and the second portion, the data set comprising a sum total of data associated with the set of selected features.

15. The computer-implemented method of claim 13 , wherein:

the forecast request is not a first request;

a new category of processed data has been added to storage since a most recent forecast, the new category of processed data comprising at least one of a new category of processed historical product data and a new category of processed historical location data; and

the forecasting method selects the machine learning model to make the forecast, selecting the machine learning model comprising:

training the plurality of machine learning models on a first portion of an expanded data set, the expanded data set comprising the new category of processed data;

validating the machine learning model on a second portion of the expanded data set; and

retraining the machine learning model on a sum total of the first portion and the second portion.

16. The computer-implemented method of claim 13 , wherein:

the forecast request is not a first request;

an additional amount of processed data has been added to storage since a most recent forecast, the additional amount of processed data comprising at least one of new processed historical product data and new processed historical location data;

the additional amount of processed data exceeds a threshold; and

the forecasting method selects the machine learning model to make the forecast; selecting the machine learning model comprising:

training the plurality of machine learning models on a first portion of an expanded data set, the expanded data set comprising the additional amount of processed data;

validating the machine learning model on a second portion of the expanded data set; and

retraining the machine learning model on a sum of the first portion and the second portion.

17. The computer-implemented method of claim 13 , wherein the forecast request is not a first request, and the method further comprises:

evaluating, by the processor, a forecast accuracy of the forecast against incoming processed historical product data; and

selecting, by the processor, the machine learning model if the forecast accuracy falls below a threshold, selecting the machine learning model comprising:

training the plurality of machine learning models on a first portion of an expanded data set, the expanded data set comprising the incoming processed historical product data;

validating the machine learning model on a second portion of the expanded data set; and

retraining the machine learning model on a sum total of the first portion and the second portion.

18. The computer-implemented method of claim 13 , wherein:

the forecast request is not a first request; and

the method further comprises:

evaluating, by the processor, a time interval between a most recent forecast and the request; and

retrain the previously selected machine learning model if the time interval exceeds a threshold, retraining comprising retraining the previously-selected machine learning model on an expanded data set comprising new processed data collected during the time interval.