Systems and methods for deriving leading indicators of economic activity using predictive analytics

Abstract

Predictive analytics techniques are used to produce leading indicators of economic activity based on factors determined from a range of available data sources, such as public and/or private transportation data. A fee-based subscription system may be provided for the sharing of leading indicators to users. A consistent, semantic metadata structure is described as well as a hypothesis generating and testing system capable of generating predictive analytics models in a non-supervised or partially supervised mode.

1 . A predictive analytics system comprising:
 a plurality of datasets received from a corresponding plurality of data sources, wherein at least one of the plurality of datasets includes first sensor data, including imaging data, derived from direct observation of activity within an environment:   a data preparation module, including a processing device, configured to assign metadata to the stored datasets;   at least one predictive analytics module, including a processing device, configured to train first and second machine learning models using the stored datasets as a source of predictor variables based on the assigned metadata for each of the machine learning models, and to use historical information regarding a metric of economic activity as a target variable, wherein the first and second machine learning models differ in at least one of machine learning model type and assigned metadata;   a hypothesis generating and testing system (HGTS) that trains the first and second machine learning models, selects the predictor variables for the machine learning models, and determines the leading indicator using the assigned metadata for the machine learning models;   the HGTS including an analytics engine configured to: assign first and second statistical metrics to the first and second machine learning models; store the assigned metadata, the first machine learning model, and the predictor variables in a primary database in response to determining that the statistical metric assigned to the first machine learning model meets a predetermined level; and store the assigned metadata, the second machine learning model, and the predictor variables in a second, probationary database in response to determining that the second machine learning model exhibits a statistical metric does not meet a predetermined level; and select between the first and second machine learning models based on their respective statistical metric; and   a publishing module configured to provide, to one or more subscribers, selected model information and output data including a leading indicator of the target variable based on the selected machine learning model and contemporaneous information received from at least one of the data sources.   
     
     
         2 . The predictive analytics system of  claim 1 , wherein the first sensor data includes at least one of audio data, imaging data, transportation equipment data, and location services data documenting transportation activity in the environment selected from the group consisting of: (a) type and activity of aircraft; (b) type and activity of vehicles, (c) type and activity of marine vessels, and (d) type and activity of trains; further wherein the metadata includes at least: data collection methods for the first sensor data; data preparation methods for the first sensor data; and at least one of accuracy, precision, or resolution of the first sensor data. 
     
     
         3 . The predictive analytics system of  claim 2 , wherein: the transportation activity corresponds to the type and behavior of at least one of aircraft, trains or marine vessels in the environment; each of the machine learning models is configured to perform object detection and classification with respect to the aircraft, train or marine vessels; and the predictor variables include the respective types and motions of the aircraft, train or marine vessels. 
     
     
         4 . The predictive analytics system of  claim 3 , further including determining and classifying the number and type of cargo held by at least one of the aircraft, train or marine vessels. 
     
     
         5 . The predictive analytics system of  claim 3 , further including determining the identity of at least one of the aircraft, train or marine vessels based on one or more visual characteristics of that aircraft, train or vessel. 
     
     
         4 . The predictive analytics system of  claim 3 , wherein: the plurality of datasets includes at least one of satellite and aerial images of the environment; and each of the machine learning models is further configured to determine at least one of the acceleration, velocity, cargo volume, cargo type, and activity over time of the aircraft, train or marine vessels based on the at least one of satellite and aerial images. 
     
     
         5 . The predictive analytics system of  claim 3 , wherein: the plurality of datasets includes elevation view images of the environment; and each of the machine learning models is further configured to determine a cargo weight for the marine vessels based on the vertical position of the vessels relative to a waterline. 
     
     
         6 . The predictive analytics system of  claim 3 , wherein: the plurality of datasets includes at least one of audio data, image data, transportation equipment data, cargo monitoring data, location services data of the aircraft, train or marine vessel or the cargo on the aircraft, train or marine vessel; and each of the machine learning models is further configured to provide real time and historical information on at least one of the aircraft, train or marine vessels or the cargo held by at least one of the aircraft, train or marine vessels. 
     
     
         7 . The predictive analytics system of  claim 2 , wherein: the transportation activity corresponds to the type and behavior of vehicles on a section of roadway; each of the machine learning models is configured to perform object detection and classification with respect to the vehicles, the occupants of the vehicles or the cargo of the vehicles; and the predictor variables include the respective types, motions, activity of the vehicles over time. 
     
     
         8 . The predictive analytics system of  claim 7 , wherein the data source associated with observation of the transportation activity is a mobile or stationary traffic camera or other imaging device. 
     
     
         9 . The predictive analytics system of  claim 7 , wherein the data source associated with observation of the transportation activity is at least one of satellite or aerial camera or other imaging device. 
     
     
         10 . The predictive analytics system of  claim 7 , wherein the plurality of data sets includes at least one of location services data, transportation equipment data of the vehicle or the cargo being carried on the vehicle. 
     
     
         11 . The predictive analytics system of  claim 2 , wherein: the transportation activity corresponds to the type and behavior of automotive vehicles parked within a region defined by at least one of a parking lot and a loading area adjacent a business, and a storage area; the machine learning model is configured to perform object detection and classification with respect to the automotive vehicles; and the predictor variables include the respective types and number motions and activity of automotive vehicles, the occupants or cargo of the automotive vehicles detected within the region over time. 
     
     
         12 . The predictive analytics system of  claim 1 , wherein at least one of the datasets includes information documenting agricultural, mining, or harvesting activity. 
     
     
         13 . The predictive analytics system of  claim 12 , wherein the datasets are selected from the group consisting of: time-lapse images comprising at least one of satellite and aerial images of one or more farms, mines, forests, pastures, storage or processing facilities; the machine learning model is configured to perform object detection and classification with respect to at least one of the vehicles, machines or equipment used, the number and type of livestock, the volume and type of crops, the volume and type of material being mined, the volume and type of lumber being harvested. 
     
     
         14 . The predictive analytics system of  claim 12 , wherein the plurality of datasets includes information documenting agricultural crop-yield data; regional climate data;
 agricultural, mining, lumber commodity market prices; field images, pasture images, forest images, crop images, crop yield/productivity data; logging yield/productivity data; herding yield/productivity data; and mining yield/productivity data.   
     
     
         15 . The predictive analytics system of  claim 1 , wherein the statistical metric is based on compiling and rank-ordering data correlations for the first and second machine learning models and determining statistical significance of the data correlations. 
     
     
         16 . The predictive analytics system of  claim 1 , wherein the plurality of datasets includes at least one of location services and vehicle, occupant, pedestrian, worker or cargo activity logs. 
     
     
         17 . A computer-implemented method for hypothesis generating and testing, comprising the steps of:
 training first and second machine learning models having corresponding first and second assigned metadata, wherein the first and second machine learning models are trained using a plurality of datasets received from a corresponding plurality of data sources, wherein at least one of the plurality of datasets includes first sensor data, including imaging data, derived from direct observation of activity within an environment:   selecting first and second predictor variables, respectively, for the first and second machine learning models;   determining a leading indicator of economic activity using the assigned metadata for the first and second machine learning models;   assigning first and second statistical metrics to the first and second machine learning models;   storing the first assigned metadata, the first machine learning model, and the predictor variables in a primary database in response to determining that the statistical metric assigned to the first machine learning model meets a predetermined level;   storing the second assigned metadata, the second machine learning model, and the second predictor variables in a second, probationary database in response to determining that the second machine learning model exhibits a statistical metric that does not meet the predetermined level; and   selecting between the first and second machine learning models based on their respective statistical metrics.   
     
     
         18 . The method of  claim 17 , wherein the statistical metric is based on compiling and rank-ordering data correlations for the first and second machine learning models based on statistical significance of the data correlations. 
     
     
         19 . The method of  claim 17 , wherein the plurality of datasets includes at least one of location services and vehicle, occupant, pedestrian, worker or cargo activity logs. 
     
     
         20 . The method of  claim 17 , wherein the plurality of datasets includes at least one of public or private economic datasets for a business, sector, local, regional or global nature.