Apparatuses and methods for evaluation of proffered machine intelligence in predictive modelling using cryptographic token staking

1. A processor-implemented method, comprising:

generating, at a host system, first data including encrypted data that does not include any information that is proprietary to the host system;

sending the first data from the host system to each data source compute node of a plurality of data source compute nodes, in response to download requests received from each data source compute node of the plurality of data source compute nodes;

receiving, via an application programming interface (“API”) of the host system and from each data source compute node of the plurality of data source compute nodes, an estimate based on the first data;

receiving an indication of a stake associated with each data source compute node of the plurality of data source compute nodes;

storing, in memory, an indication of a predefined feedback resource associated with a smart contract;

ranking each data source compute node of the plurality of data source compute nodes based on the received indications of stakes, to generate a plurality of ranked data source compute nodes;

calculating, via at least one processor of the host system, an accuracy of each received estimate by comparing the received estimates to second data;

for a first ranked data source compute node from the plurality of ranked data source compute nodes:

decrementing the predefined feedback resource and assigning a token augmentation to the first ranked data source compute node if the accuracy of the estimate associated with the first ranked data source compute node has a log loss of less than about −ln(0.5); and

for each remaining ranked data source compute node from the plurality of ranked data source compute nodes:

determining a value associated with the predefined feedback resource;

if the value is greater than zero, decrementing the predefined feedback resource and assigning a token augmentation to the ranked data source compute node if the accuracy of the estimate associated with the ranked data source compute node exceeds a predefined threshold.

2. A method, comprising:

sending first data to each data source compute node of a plurality of data source compute nodes;

receiving, from each data source compute node of the plurality of data source compute nodes, an estimate based on the first data;

receiving an indication of a stake associated with each data source compute node of the plurality of data source compute nodes;

storing, in memory, an indication of a predefined resource;

ranking each data source compute node of the plurality of data source compute nodes based on the received indications of stakes, to generate a plurality of ranked data source compute nodes;

calculating, via a processor, an accuracy of each received estimate by comparing the received estimates to second data;

for a first ranked data source compute node from the plurality of ranked data source compute nodes:

decrementing the predefined resource and assigning a token augmentation to the first ranked data source compute node if the accuracy of the estimate associated with the first ranked data source compute node exceeds a predefined threshold; and

for each remaining ranked data source compute node from the plurality of ranked data source compute nodes:

determining a value associated with the predefined resource;

if the value is greater than zero, decrementing the predefined resource and assigning a token augmentation to the ranked data source compute node if the accuracy of the estimate associated with the ranked data source compute node exceeds the predefined threshold.

3. The method of claim 2 , further comprising receiving, from each data source compute node of the plurality of data source compute nodes, an indication of a confidence level wherein the ranking each data source compute node of the plurality of data source compute nodes is further based on the received indications of confidence level.

4. The method of claim 2 , wherein the predefined resource is defined by a smart contract, and decrementing the predefined resource includes one of modifying and replacing the smart contract.

5. The method of claim 2 , wherein receiving the estimates from each data source compute node of the plurality of data source compute nodes is via an application programming interface (“API”).

6. The method of claim 2 , wherein the estimate associated with the first ranked data source compute node exceeds a predefined threshold if log loss<−ln(0.5).

7. The method of claim 2 , further comprising calculating, via the processor, a consistency of each received estimate,

wherein decrementing the predefined resource and assigning the token augmentation to the first ranked data source compute node is further predicated on the estimate associated with the first ranked data source compute node having a consistency of at least 75%.

8. A method, comprising:

sending first data from a host node to each data source compute node of a plurality of data source compute nodes, the first data including target data;

receiving, at the host node, an indication of a stake associated with each data source compute node of the plurality of data source compute nodes;

sending, in response to receiving the indications of stake, a stake identifier to each data source compute node of the plurality of data source compute nodes;

receiving, at the host node and from each data source compute node of the plurality of data source compute nodes, a target estimate based on the first data and associated with the target data; and

calculating, at the host node and for each received target estimate, a predictive accuracy value associated with the target estimate;

storing the calculated predictive accuracy values associated with the target estimates;

for a first data source compute node of the plurality of data source compute nodes:

decrementing a predefined divisible resource and assigning a value augmentation to the first data source compute node if the predictive accuracy value associated with the first data source compute node exceeds a predefined threshold; and

for each remaining data source compute node of the plurality of data source compute nodes:

determining a value associated with the predefined divisible resource, and

if the value associated with the predefined divisible resource is greater than zero, decrementing the predefined divisible resource and assigning a value to the data source compute node if the predictive accuracy value associated with the data source compute node exceeds the predefined threshold.

9. The method of claim 8 , wherein the first data includes target data, feature data, and identifier data.

10. The method of claim 9 , wherein the data further includes at least one of: era data or indications of data type.

11. The method of claim 8 , wherein the sending the first data to a first data source compute node of the plurality of data source compute nodes is in response to a download request received from the first data source compute node.

12. The method of claim 8 , wherein the sending the first data is via an application programming interface (“API”).

13. The method of claim 8 , further comprising calculating, at the host node and for each received target estimate, a consistency value associated with the target estimate.

14. The method of claim 8 , further comprising calculating, at the host node and for each received target estimate, an originality value associated with the target estimate.

15. The method of claim 8 , further comprising calculating, at the host node and for each received target estimate, a concordance value associated with the target estimate.

16. The method of claim 8 , further comprising:

calculating, at the host node and for each received target estimate, a consistency value associated with the target estimate, an originality value associated with the target estimate and a concordance value associated with the target estimate; and

sending a signal to cause the calculated predictive accuracy values, the consistency values, and the originality values to be viewable to each data source compute node of the plurality of data source compute nodes.

17. The method of claim 8 , further comprising:

sending, in response to receiving the indications of stake, and to each data source compute node of the plurality of data source compute nodes, at least one of: a stake status, a transaction hash, or an indication of a stake value.

18. A method, comprising:

storing, in memory, an indication of a predefined divisible resource;

ranking each data source compute node of a plurality of data source compute nodes based on indications of stakes associated with each data source compute node of the plurality of data source compute nodes, to generate a plurality of ranked data source compute nodes;

calculating, via a processor and for each data source compute node of the plurality of data source compute nodes, an accuracy of a target probability estimate received from the data source compute node by comparing the received target probability estimate to a stored dataset;

for a first ranked data source compute node of the plurality of ranked data source compute nodes:

decrementing the predefined divisible resource and assigning a value augmentation to the first ranked data source compute node if the accuracy of the target probability estimate associated with the first ranked data source compute node exceeds a predefined threshold; and

for each remaining ranked data source compute node of the plurality of ranked data source compute nodes:

determining a value associated with the predefined divisible resource;

if the value is greater than zero, decrementing the predefined divisible resource and assigning a value augmentation to the ranked data source compute node if the accuracy of the target probability estimate associated with the ranked data source compute node exceeds the predefined threshold.

19. The method of claim 18 , wherein the first ranked data source compute node exceeds the predefined threshold if log loss<−ln(0.5).

20. The method of claim 18 , further comprising:

calculating, for a data source compute node of the plurality of data source compute nodes, a consistency value associated with the target probability estimate received from that data source compute node; and

excluding that data source compute node from the plurality of ranked data source compute nodes if the consistency value is below 75%.

21. The method of claim 18 , wherein ranking the data source compute nodes is further based on indications of confidence received from each data source compute node of the plurality of data source compute nodes.

22. The method of claim 18 , wherein the value augmentation has a value of s/c, where “s” is associated with the indication of stake, and “c” is associated with the indication of confidence.